Exploring the Applicability of Production Sharing Agreements to Offshore Wind Projects

EXPLORING THE APPLICABILITY OF PRODUCTION SHARING

AGREEMENTS TO OFFSHORE WIND PROJECTS

Chapter

1introduction

The purpose of this exploratory research is to examine whether hydrocarbon

PSA-like contract terms could be applied to OWE projects and is the first to attempt

such analysis. PSAs are used in the hydrocarbon industry in more than eighty

countries, most of which are considered developing or undeveloped nations (Wood

Mackenzie, 2022). The contract governs the terms of the lease of seabed for O&G

exploration, development, and production (Johnston, 1994). It outlines the rights,

duties, and obligations of the contracting parties. The lessor in the agreement, the

party who leases the seabed for the hydrocarbon project, is the resource owner,

which in most countries globally is the government (hereinafter known as the

Government). The lessee is typically the energy company or investor who will be

building and operating the projects (hereinafter known as the Contractor). PSAs are

widely used in the hydrocarbon industry (Kaiser & Pulispher, 2005a) and when

properly structured,

are considered a successful and attractive contract by which governments lease seabed

to industry (Luo & Yan, 2010; Pongsiri, 2004; Zhang, 2022).

OWE has in the past required government subsidisation to be economically

attractive, but this era is ending (Jansen et al, 2022). Countries, such as developing

nations, which could not afford to subsidise offshore wind, may wish to encourage a

national energy transition. This transition could be away from polluting hydrocarbons

towards renewable energy. It could be away from energy dependence, if no

hydrocarbon industry is present, towards independence through renewables which

may be sourced locally. There is no requirement for these countries to follow existing

offshore wind leasing models, conversely, there is an opportunity for them to shape

new contract terms for offshore wind. This research looks at capturing contractual best

practices from the O&G industry and testing their applicability to offshore wind, to

explore an alternative leasing mechanism for future projects.

Successful application of the contract terms to offshore wind projects depends

upon the project having a positive financial outcome for both parties to the

agreement. A positive financial outcome is defined as the generation of a positive net

present value (NPV) for both parties. Thus, the quantitative phase of this exploratory

mixed method research addressed five research questions (RQ) through stochastic

cash flow modelling in combination with multiple regression (“meta-modelling”) and

scenario analysis. The use of these methods in combination unifies the best of

industry practices and academic rigour. These RQs concern the economic viability of

the PSA- like terms, their impact on offshore wind project profitability, the optimal

set of contract terms, and the effect of project characteristics on profitability. RQs 1, 2

and 3 (below) take into account the perspective of both parties to the contract, which

is an advance on earlier research.

RQ1-1: Is the wind farm characterised by a positive NPV for the Contractor under the

PSA contract terms modelled?

RQ1-2: Does the wind farm provide a positive NPV to the Government under the PSA

contract terms modelled?

RQ2-1: What effect does each contract term have on Contractor NPV?

RQ2-2: What effect does each contract term have on Government

NPV?

RQ3-1: What effect do other inputs such as Discount Rate, Price, and Power

Generation have on Contractor NPV?

RQ3-2: What effect do other inputs such as Discount Rate, Price, and Power

Generation have on Government NPV?

RQ4: Can the optimal set of contract terms be discovered for a given wind farm?

RQ5: How do the NPVs fare under different power generation, price, and cost

conditions?

Adoption of the contract by governments as a leasing instrument for offshore

wind is also dependent on it having attractive features not present in existing

contracts.

The qualitative phase explored the reasons behind using PSAs for hydrocarbon

leasing and the benefits which could accrue if PSA-like contract terms were used for

OWE. This phase of the research employed interviews with O&G industry

professionals to address RQs 6 and 7:

RQ6: What are the advantages and disadvantages of using PSAs in hydrocarbon

leasing?

RQ7: What could be the advantages and disadvantages of using PSA-like contract

terms in leasing for OWE projects?

A successful outcome of this research involves finding that PSA-like terms can be

used for OWFs and that there will be benefits for both contracting parties. These

factors would entice a Government to offer and a Contractor to accept the terms. In

that case, this research would provide a new type of leasing arrangement for OWE that

may attract developing countries to add another renewable source to their energy

transition.

CHAPTER 2

REVIEW OF LITERATURE

Note on Grey Literature

Grey literature is defined as publicly available writings which are not controlled

by commercial publishing (Lawrence et al, 2015) i.e., have not gone through the

traditional peer review and publishing process of an academic journal. Its

incorporation into research can be positive, and is becoming increasingly important

(Adams et al, 2017). Pappas & Williams (2011) point to the delay between research

and peer-reviewed journal articles in their field of medicine and argue that grey

literature fills the time gap between cutting-edge research or innovation and

publication. Grey literature is produced by governments, academics, businesses, and

industry; it is protected by intellectual property rights; and whilst it is often of

sufficient quality to be preserved in libraries, it is not peer reviewed (Adams et al). In

the course of this study, the researcher has observed the progression of OWE to be so

rapid, that to rely solely on academic literature would mean missing the latest trends

and events. Thus, the following literature review contains references to articles,

reports, and web pages which have been published but not peer-reviewed.

Energy Transition

“Our addiction to fossil fuels is pushing humanity to the brink. We face a stark

choice: Either we stop it — or it stops us........We are digging our own graves." - U.N.

Secretary-General Antonio Guterres at the COP26 conference in Glasgow, UK in

November 2021 (Associated Press, 2021).

As world leaders acknowledged at the 26th Conference of Parties (COP26), the

existential threat to the planet from climate change must be stemmed. Energy derived

from hydrocarbons must be significantly reduced and replaced by energy originating

from renewables (Alolo et al, 2020). Indeed, a shift from hydrocarbons is key for

reaching already agreed-upon climate goals, as well as the more ambitious goals that

will be required to halt climate change (IRENA, 2019). Many countries are well-

endowed with renewable energy resources, including solar, geothermal, hydropower,

wind, and biofuels. In 2022, 40% of global installed power capacity came from

renewable sources, which was a 5% increase in renewable capacity over the prior year

(IRENA, 2023). More startlingly it represents a 48% increase over a decade before.

Encouragingly, according to Haegel & Kurtz (2023), whose data is used by the

National Renewable Energy Laboratory (NREL), 83% of new power generation

capacity added in 2022 came from renewables. Contrast that with 2018 when the

number was only 65%. This research aims to provide one small piece of the energy

transition jigsaw and demonstrate that the use of hydrocarbon PSA-like terms could

provide a leasing alternative for OWE to developing nation governments.

Renewable Energy’s Benefits

A drive towards renewable energy can mitigate two energy-related challenges for

a nation: dependency on fossil fuels, often imported from other countries, and climate

change (Lee, 2021). Some countries, such as Jordan, which imports 94% of its energy

needs, see an opportunity for improved energy security through developing domestic

renewable energy sources (Abu-Ramman et al, 2020 and IRENA, 2019). While this

will

not mitigate the need to import liquid hydrocarbon fuels for transport, it can still have

beneficial environmental and security of supply impacts on domestic power generation.

Similarly, Namibia imports its power from neighbouring South Africa, Zimbabwe,

and Mozambique; those contracts can fill 92% of Namibia’s peak demand. However,

the country has the potential off its own shores to fulfil its demand with OWE (Choti

& Xydis, 2022) and become power independent. The same two reasons cause

Bangladesh to have ambitions to tap into its own solar and wind (initially onshore,

subsequently offshore) assets rather than import increasing amounts of hydrocarbon

fuel (Alam, 2022).

Security of energy supply and climate change are two of the four reasons driving

Russia to pursue policies to support renewable energy (Kudelin, 2021) – the other

reasons being a decrease in negative health effects and the associated cost, and

replacement of declining hydrocarbon resources. Pakistan (Rabbani et al, 2022) sees

four reasons to switch to renewable energy: energy security, reduced use of fossil

fuels, economic benefits, and job creation. Nigeria (Idris et al, 2012) has five reasons

for a switch to onshore wind energy: security of supply, decreased reliance on

hydrocarbons, fewer negative effects on the environment, increased socio-economic

development, and improved quality of life.

Vietnam does not subscribe to the common belief that pursuing green policies

will stunt development, but rather that it will accelerate due to the adoption of

technology and improved energy efficiency (Zimmer et al, 2015). The country wants

to improve energy supply security, act as a role model, and attract foreign investment

and donors due to its responsible attitude towards climate change (Zimmer et al). The

authors find

Vietnam suffers some of the worst air pollution in the world, has a long coastline

making it vulnerable to rising sea levels, and the coal, oil, and gas that fuelled the

economic development since the 1990s are being depleted. A shift to renewable

energy makes sense.

Atteridge et al (2012) discuss the shift in policy in India, which had very low

greenhouse gas emissions per capita between the 1990s and 2012. The former two-

fold policy was centred around the country not being part of the climate change

problem, and that spending on a shift to renewable energy would compromise the

national goals of poverty reduction and economic growth. The authors find that by

2012 India had become concerned with energy supply security and decreasing imports

of fossil fuels, both of which could be achieved by a shift to renewable energy. They

state that the goals of energy security, economic growth, and poverty reduction can be

achieved in a way that has a side-benefit of mitigating climate change. Thus, India is

beginning to emulate wealthy countries in some policies.

Gopalakrishnan et al (2023) suggest that the “fragile states” – often developing

countries – can achieve wider access to power at a lower cost with renewable energy

than with fossil fuels. They can benefit locally from the transition and not just through

their contribution to the global drive to mitigate climate change.

Further, the expectation (Tavakolian et al, 2022) is that hydrocarbon exploitation

will become more expensive in the future as the easy-to-access, conventional

resources are exhausted. This will make renewable energy more competitive.

Additionally, new sources, such as oil sands, have a more deleterious effect on the

environment than sources already tapped, such as light, onshore oil deposits. It is

true that the shift in

recent decades from coal- to natural gas-fired power plants is better for the

environment. In the US alone, 121 coal-fired power plants were converted to natural

gas between 2011 and 2020 (Aramayo, 2020). However, this improvement may be lost

if future sources of natural gas are sufficiently difficult to access and transport that the

harm to the environment accelerates again.

The Problem of Intermittency

Naysayers often point to the intermittent nature of renewable energy sources as a

barrier to success, and indeed this is a major concern in comparison to stable baseload

technologies using hydrocarbons as fuel. However, Bhandari (2014) and IRENA

(2019), discuss the clustering of two or more renewable technologies, known as hybrid

renewable energy systems (HRES). Such systems improve the intermittent supply issue

inherent in naturally occurring power sources. These can be connected to the existing

grid so that power can be pulled from the grid if the renewables cannot meet peak

demand, eliminating the need for storage - (Bhandari). If the system is not connected, it

can include lithium battery storage, but the battery can be smaller and cheaper than in a

single source system, whose intermittency problem will be greater (Bhandari). Some

use diesel fuel to replace renewables when demand cannot be met (Bhandari). That

offers a transition away from hydrocarbons but does not entirely replace them.

However, the HRES/diesel combination does solve the unnecessary expense of

oversizing the system for peak demand (Bhandari). The author, in his 2014 literature

review, points to studies, some theoretical and some functioning in practice, designed

as follows: solar and wind with battery; solar and wind with grid connection; solar and

hydro with grid connection; solar and wind with diesel (seen in Palestine); solar, wind

and hydro with

battery; solar, wind, battery and diesel (seen in Malaysia); and wind-pumped hydro in

Ethiopia (the water, pumped to a height acts as the storage mechanism). These projects

are all small-scale, but solutions such as these are expected to be seen increasingly,

and since Bhandari’s 2014 paper, hydrogen has emerged as a new storage technology

(IRENA, 2019).

In 2017, a larger-scale hybrid project of 43 MW of onshore wind, 15 MW of solar

and a relatively small, 2 MW, battery was installed in Australia (IRENA, 2019). In

2018 the first HRES including offshore wind became operational when the 30 MW

Hywind OWF in Scotland was connected to a battery array (IRENA, 2019). The

hybrid system, named Batwind, decreases volatility in power supply by storing energy

in a substation of 1 MW batteries (Equinor, 2018). The world’s first OWE and solar

project is currently under development off Nanri Island, China (Memija, 2023b). This

project comprises 12 MW of offshore wind on three floating semi-submersible

structures which also house solar panels. In addition, fish farms could also be set up in

the centre of each structure. Larger-scale HRES involving offshore wind will require

more advances in technology such as floating solar or wave power. Pilot and

conceptual projects are being undertaken by Scandinavian, British, German, Spanish,

and American companies (Solomin, 2021). Rusu et al (2017) point to the opportunity

for offshore wind HRES in developing countries (OWE in combination with wave

energy), though at the time of writing, the authors found that the cost would have been

prohibitively high.

The adoption of any new energy source will have problems, as does the

acceptance of a new technology, a new medical procedure, or any advance. Human

innovation has a history of overcoming such issues. For example, the use of oxen

was superseded by

more powerful horses in the ninth century, but that transition was not possible until the

horseshoe was invented (Fouquet, 2010). The same author writes of the sixteenth-

century introduction of chimneys and grates ushering in the switch from wood to coal

for heating. He also notes that new technologies accompanied the transition to fossil

fuels, first seen in the UK. From there the technologies spread globally with the shift

in fuel type. Innovation will continue to be seen to accompany the transfer to

renewable energy sources.

This research aims to tap into the benefits of renewable energy, applying a widely

used practice from O&G, giving renewable energy a boost up the learning and

innovating curve with respect to offshore leasing.

Adoption of Renewable Energy

Nations such as those in the G20 (mostly developed and wealthy) accounted for

85% of global GDP and 66% of global population in 2022 (OECD, 2022). In 2023,

they held 87% of installed renewable energy capacity (calculated from IRENA, 2023),

leaving only 13% for non-G20 (mostly developing) nations. Within the IRENA G20

capacity data, the G20 countries which are considered developed (this includes China)

account for 84% of installed renewable energy capacity. Developing G20 nations only

account for 16%. This focus on developed countries is due to the historical costs

involved in developing a renewable energy industry – costs over and above the

expenditure required to provide energy from fossil fuels at the time. The dominance of

wealthy countries in renewable energy is also due to the need for research and

development (R&D) to generate the technologies needed – expenditure which is out of

reach of many developing countries. E&Y (2021) identify that developed countries

must take responsibility for both financing R&D for renewables and assisting

developing countries with financing for renewable energy projects. Bui et al (2022)

point specifically to Vietnam’s potential for OWE but that the country will rely on

technology flow from developed nations. Do et al (2021) speak of Vietnam’s likely

reliance on knowledge and manufacturing from other countries.

Although developing countries lag behind the progress of their wealthy

neighbours, they stand to benefit greatly from renewable energy as they develop,

rather than relying on hydrocarbons. Industrialisation, increased trade, urbanisation,

deforestation, and combatting poverty in developing countries are increasing

contributors to environmental degradation when fueled by hydrocarbons (Masron &

Subramaniam, 2021). An acceleration of renewable energy penetration into those

economies is essential so that those countries do not become the next generation’s

biggest polluters, affecting both local and global environments. For example, in

Africa, the second most populous region in the world (Wittgenstein, 2022), 80% of the

rural population still relies upon candles and firewood for lighting and cooking

(Nalule et al, 2022). Replacement of wood for cooking in Africa is expected to save

US$20-30 billion a year in reducing health problems caused by poor indoor air quality

(Gopalakrishnan). In turn, Cottrell & Falcao (2018) point out that developing and

undeveloped countries are the hardest hit by climate change and stand to benefit from

the energy transition. Welsby et al (2021) indicate that “60% of the oil and fossil

methane gas, and 90% of coal must remain unextracted to keep within a 1.5 deg C

carbon budget”. Yet, between 1999 and 2019, O&G production increased by 12% and

93% respectively in Africa (BP, 2021 & BP, 2010).

Despite the potential cost, but recognising that renewable energy is beneficial and

that the country is more exposed to climate change than most, Vietnam has

implemented at least seven policies to encourage renewables or discourage

hydrocarbons (Zimmer et al). As Vietnam is a developing country, it is not bound by

obligations under the Kyoto Protocol, yet it created goals to reduce greenhouse gases

by 2020 and beyond, and policies around carbon tax, emission reduction, green

certificate trading, increasing energy efficiency, phasing out fossil fuel subsidies, and

increasing tax on hydrocarbons (Zimmer et al). By 2022, these had evolved into goals

for emission reductions in 2030 and 2040, 75% power from renewables by 2045, the

phase-out of coal by the 2040s, and to be net-zero by 2050 (Agarwal et al, 2022). In

2022, Vietnam had the third largest renewable energy capacity in Asia, behind much

larger China and India (IRENA, 2023).

Wind energy is seen as having advantages over other renewable energy sources

including ease of implementation, long life of turbines, uncomplicated infrastructure,

and being low-cost (Rabbani et al). As the International Energy Agency (IEA, 2019)

points out, wind energy, and particularly OWE, has less variability than competing

renewable energy sources such as solar. Tidal, wave, and geothermal power have the

advantage over wind of being available around the clock. However, geothermal is

currently cost-prohibitive except in niche settings (Tester et al, 2021) and tidal energy

demonstrations have also suffered the same fate of being too expensive (Chowdry et

al, 2021). Wave power has seen many successful small-scale demonstrations but there

is not yet a commercial-scale park due to technological constraints, prohibitive capital

costs, and high operating costs (Guo & Ringwood, 2021).

Compared to fossil fuels, renewable energy is a new industry even in countries

with the longest histories of renewable energy installations. The most developed

example is biofuel ethanol, with a forty-year history since the early 1980s (Cotti &

Skidmore, 2010). In contrast, OWE is one of the newest, in which installed capacity

only reached double-digit MW output at the turn of the twenty-first century (Higgins

& Foley, 2014). Developments and changes to all aspects of the industry are frequent

and common, as in any growth area, providing opportunities for innovation, not only

in the technological and engineering sense but in all facets of the business. As of the

late 2010s, OWE has shifted from being considered a new technology to a “rapidly

maturing” one (IEA, 2019 p15). Nowadays, the new technology related to offshore

wind is floating offshore wind. It has only been considered feasible since 2017 with

the 30 MW Hywind demonstration (pilot) project in Scotland (IEA, 2019). By 2023

Hywind had been joined by three more: 25MW Windfloat Atlantic off Portugal,

50MW Kincardine off Scotland, and 88MW Hywind Tampen, off Norway (Edwards

et al, 2023). In December 2022, construction began in China on what will be the

world’s largest floating wind farm, which at 1GW will have eleven times the capacity

of Hywind Tampen (Xia et al, 2023). Floating OWE is considered to have even more

potential than fixed bottom OWE due to higher wind potential and lower visual and

acoustic effects on human populations further offshore (Barooni, 2023). The authors

point out that since much of the usable wind is located where the seabed is over 60m

deep and/or the bathymetry is complex, floating solutions are required. Constant et al

(2021) confirm this, approximating that 80% of offshore wind sites are at water

depths greater than 60m and Casey (2024) states 70% of USA OWE is only accessible

with

floating installations. Barooni et al detail the five types of floating structure that can

be used – interestingly, four of these are taken directly from offshore O&G platform

designs. However, as those authors admit, the costs of floating OWE are higher than

fixed bottom, which are in turn higher than onshore.

This research, with its study of the applicability of an existing lease contract

type, aims to encourage the adoption of one renewable energy source; fixed bottom

OWE. If successful, the applicability to other sources could be demonstrated in future

work.

Offshore Wind

Introduction

Wind energy, and specifically OWE, have enormous potential to provide power

for the global population – eighteen times over - according to the IEA (Yale

Environment 360, 2019). Indeed, the best sites alone could supply more power than

the world consumed in 2019 (IEA, 2019). Wind, together with solar power, will need

to become a central pillar of global energy supply by 2050 if the world is to meet the

net zero roadmap set out by the IEA (GWEC, 2022). Wind will provide 35% of global

power and solar 33%. This includes moving power generation from wind offshore;

there is more available space offshore compared to on land (Do et al, Esteban et al,

2011, Lee et al, 2023). The wind is stronger out to sea (Dahl, 2022), and wind speed

is more constant offshore (Esteban et al). Bigger turbines and farms can be installed

offshore without creating conflict for land use (Esteban et al). As IRENA (2019)

points out, Gigawatt-scale farms can be built offshore, close to population centres. To

achieve the same scale onshore would require over 300 turbines rated at 2.75 MW

each (USGS, 2023) vs. 67 offshore turbines of 15 MW each (calculated from Lewis,

2023b) or 56 offshore turbines of 18 MW, prototyped in 2023 (Lewis 2023a). Taller

turbines can be built offshore, which will have a smaller visual and acoustic impact

from land (Soares-Ramos et al, 2020) and are expected to increase the efficiency of

wind energy (Chen & Su, 2022). Visual and noise impacts are two of society’s

objections to onshore wind farms, as are decreases in property values, destruction of

views, blocking of light, and diminished access to nature (Westlund & Wilhelmsson,

2021). All these concerns can be mitigated, or addressed entirely, by moving wind

farms offshore. However, Dwyer and Bidwell (2019) caution that local communities

must still be properly engaged for offshore installations. In fact, tourists to Block

Island, offshore Rhode Island, USA, view the wind farm positively and are interested

in seeing it (Bidwell, 2022). Residents, when surveyed and interviewed before

installation, at first operations, and a year later, became more positively disposed to

the OWF (Bingaman et al, 2023).

In their bibliometric analysis, Chen & Su find OWE has both positive and

negative impacts on marine life populations. On the negative side, transportation,

piling, seabed disturbance, and noise can cause sea life to avoid the area. However,

once in place, on the positive side, the turbine bases can act as artificial reefs and

attract sea life. This has long been seen around O&G rigs and facilities (Ajemian et al

2015). To achieve the global temperature-rise limitation of 1.5 degrees Celsius,

onshore wind will need to scale up by a factor of three by 2030 and a factor of nine

by 2050 (IRENA, 2019). Offshore wind, being less prevalent today, has further to go:

a ten-fold increase by 2030 and over forty times by 2050.

Offshore Wind Usage and Policy Today

In 2022, according to IRENA (2023), OWE capacity was split 51:49 between

Asia and Europe. Within Europe, the split is 55% EU to 45% UK, with the UK

holding almost 70% more capacity than the next individual country, Germany. In

Asia, China holds 94% of the capacity, with more than double the UK’s. The US lags

with 0.07% of the global OWE share though in 2024 (BOEM) there are thirty-nine

active federal offshore leases at various stages of analysis, permitting or development.

In 2022, the first auction of seabed in four years was completed and garnered record

bids of US$4.4bn, ten times that of prior auctions; and encouragingly, more than

recent O&G leasing auctions (Lewis, 2022). The 2024 five-year offshore wind leasing

schedule (BOEM) envisages twelve more lease sales to continue to support the goals

of 30GW of fixed bottom offshore wind by 2030 and 15 GW of floating offshore

wind by 2035. Being a follower into a new technology may even prove to be prudent.

Hayashi et al (2018) opined that China has rushed the building of wind generation

capacity and has neglected to build a strong foundation including quality

infrastructure, knowledge, and a skilled workforce for the future. Further, as

Westwood (2022) points out, China will be challenged to add capacity at the same

rate after the expiration of governmental subsidies for offshore wind at the end of

2021.

IRENA (2023) figures show that offshore wind capacity additions increased 29% in

2019, 52% in 2020, and a huge 194% in 2021 to beat the expiry. The same data shows

a modest, but non-zero capacity addition in 2022 of 15% which matches the average

rate of growth in onshore wind from 2016 to 2022.

Taiwan offers an example of a country which realises the key role offshore wind

can play in the energy transition (Yue et al, 2021). Today, power generation is from

hydrocarbons (coal and natural gas), 98% of which are imported, and nuclear. The

Government underwent a policy shift in 2016 towards nuclear and natural gas, but the

nuclear expansion has halted since the dangers were realised post-Fukushima (Yue et

al). Now the Taiwanese Government is shifting towards offshore wind, realising the

strong offshore potential, and has specified thirty-one OWFs that will start to replace

the other energy sources (Yue et al). The authors demonstrate that nuclear and coal

can be replaced by OWE by 2025 and 2032 respectively, and natural gas can be

replaced by 2040. Similarly, Hong Kong, which imports most of its hydrocarbon fuel

for power from mainland China, sees an opportunity to decrease its carbon footprint

and achieve supply resilience by taking advantage of the good wind potential in the

surrounding waters (Delina, 2022).

France is an interesting case, since 80% of its power comes from nuclear energy,

15% from hydro, and very little from fossil fuels (Desvalles et al, 2023). Other forms

of energy, including offshore wind, are not welcomed due to the large footprint of the

installation per unit of power produced. The authors point out that conversely, France

does not reject the concept of OWE, and the country provides one-third of Europe’s

manufacturing capacity for turbines. However, most stakeholders do not see a place

for offshore wind in France’s energy mix.

Most countries will also face challenges from aging or insufficient grid

infrastructure e.g., Vietnam, China, Germany (Do et al), and the USA (Smith et al,

2019) though if there is the will, there is time to rectify this issue alongside the

building of large offshore wind installations. In 2022 (IRENA, 2023) offshore wind

comprised 7% of total wind energy and less than 2% of renewable energy – there is

time to address infrastructure issues. This author is not attempting to minimise this

critical issue, but to leave its resolution to others.

There are currently very few developing countries engaged in offshore wind

activity, due to the historical costs and requirements for subsidisation. Vietnam is the

only example of modest active offshore wind with small coastal power plants serving

as pilot projects to test the area (Tsuge et al, 2023). The report does highlight that the

testing has been successful and there are plans for a 500 MW to 1 GW farm by 2030.

2023 also sees the first leasing of four tracts of seabed off India (Times of India, 2023)

for 4 GW of power generation potential. The article mentions ten further tracts to be

auctioned in the future for another 10 GW of generation. This adds to the 42 GW of

onshore wind currently active, and 58 GW planned by 2030 – to make up nearly a

quarter of India’s ambition for 500 GW of renewable energy by 2030. Brazil is leasing

very successfully for offshore wind, having received 71 applications for projects by

late 2022 with a total proposed capacity of 177 GW, including large 2 GW and 3 GW

sites (Buljan, 2023).

None of these leasing efforts use a contract modelled on a PSA. This research

aims to widen the options for governments wishing to lease seabed for offshore wind

projects.

Global Wind Potential

OWE is one of the most rapidly developing renewable energy sources (Dahl et al,

2022). Consider that in 1991, the first OWF became operational in Denmark, with a

capacity of merely 5 MW of power (Gaudiosi, 1996). In 2022 OWE capacity was over

twelve thousand times that, at 63,200 MW (IRENA, 2023). One hundred OWFs have

been put in place between 2000 and 2022, and a thousand more are in the development

stages (Chen & Su).

The strongest resources in terms of wind potential are to be found in the

temperate zones offshore Northwest Europe, Northern USA and Canada, China,

Southern Australasia, Southern and Northern Latin America, Southern Africa (Weiss,

2018); Northwest and Northeast Africa (Soares et al, 2020); and Eastern Africa

(Choti & Xydis, 2022). The best geographic locations can be seen in Figures 1 and 2

below, in which the highest potential areas are shaded in yellow/green (Figure 1) and

yellow- orange-red (Figure 2):

Figure 1: OWE Potential. Source: Soares et al (2020)

Figure 2: OWE Potential. Source: Weiss et al (2018)

Since only two of those areas, Northwest Europe and China, have any significant

OWE capacity to date, there is much scope for expansion. Eicke et al (2022) posit that

by 2050 60% of offshore wind capacity will be sited in Asia, 22% in Europe, and 16%

in North America. This leaves little for Africa and Latin America, which both have

wind potential and in the case of Africa, a large population. The continent is predicted

to hold 25% of the world’s population by 2050 (Wittgenstein, 2022). The World Bank

(2019a) also recognises in its programme the opportunity to fast-track the adoption of

OWE in developing countries. However, the programme is about raising awareness,

assessing potential, and identifying challenges rather than installing wind farms.

The closer wind farms are to population centres the better, as power attenuates

when transported (IEA, 2014), and the cost to build or upgrade infrastructure is lower.

There is a calculable tradeoff between placing wind farms in the areas of high

potential that are further from the markets and siting them close to large cities in

lower wind potential areas (Lamy et al, 2016). The authors state that it is important to

study

this trade-off – they demonstrate that in the mid-west of the USA, moving to an area

with a wind capacity of 43% vs. 40%, extra infrastructure capital expenditure of

US$1.4 billion can be made before reaching the breakeven point. The authors also

point to the relevance of these calculations in China, where the best wind potential is

in the Northwest, but the demand is in the East. This argument can be extrapolated to

any country or continent with higher wind potential far from the demand centres.

Lamy et al were concerned with onshore wind power but the point is also relevant to

offshore wind. 84% of countries have a coastline (Rusu et al), and 41% of the global

population live within 100km of a coast. In over half of the world’s coastal

countries, 80-100% of their population lives within 100km of the coast (Martinez et

al, 2007). There is a large market for power from offshore wind.

A diverse set of organisations see a strong future for wind power’s penetration

into the global energy mix (IRENA, 2019). Agencies such as IRENA, the IEA, the

consultancy BNEF, energy companies such as Equinor, DNV, and Shell, and the

NGO Greenpeace see wind providing 21-35% of global power by 2050 under current

policies (IRENA, 2019). This share will be even higher if policies are amended to

limit global temperature increase to 1.5 degrees Celsius.

Learning from the Oil and Gas Industry

The offshore wind industry has much in common with the offshore O&G industry

such that offshore wind can make quick advances by emulating or using O&G

resources. Edwards & Dalry (2011) find six common technologies; including offshore

construction vessels, remote-operated vehicles, and dynamic positioning systems, are

directly relevant and transferable to offshore wind operations. The authors also found

in their interviews that the supply chain has significant similarities, but they caution

against one O&G practice, namely sub-contracting for servicing and maintenance, due

to the risks introduced by not maintaining control of operations. Finally, they advocate

for the adoption of one practice; cooperation on technology, research, and

development, which has been a key success factor in O&G. At the time of writing,

offshore wind technologies had been developed in secret, and opportunities to

collaborate were being lost.

O&G companies have recognised the similarities, and many have entered the

OWE space. Mathews et al (2022) point to TotalEnergies of France, BP of the UK,

Equinor of Norway, and CNOOC of China – all global major O&G companies, who

have either entered joint ventures to build OWFs or have started their own OWE

divisions. In addition, service companies are transitioning to gain offshore wind

capabilities; thus, all these companies are engaging in Schumpeter’s creative

destruction process and replacing their own O&G business with renewable energy

(Mathews et al).

It is due to these similarities that offshore wind has been selected as the

renewable energy source for this research and as a topic of study for the application

of PSA-like terms. OWFs are the most similar renewable energy installation to

offshore O&G platforms, to which PSAs currently apply. This researcher does not try

to claim that OWE can replace O&G – merely that offshore wind can provide power

where traditionally hydrocarbons would be used – and that best practices from the

O&G industry could potentially apply in leasing contract terms.

Hydrocarbon Contract Terms

Global hydrocarbon production is governed by fiscal systems which are intended

to lay out the contractual rights and obligations between the foreign investor (known

as the Contractor) and the host Government; and provide the fiscal calculations for

the allocation of revenue (Johnston). In this sense, fiscal refers to streams of

government revenue; these could be traditional taxes or non-tax cash payments

(Sunley, 2003).

There are two primary fiscal systems, the Concession Agreement (CA) and the PSA

(Løvås & Osmundsen, 2009). Other less-used systems include a standalone strategy in

which foreign investment is not sought (Mohamed et al, 2022a), and Service

Contracts, which are unpopular with foreign investors who merely receive a fee for

the work and do not share in the upside; nor does the system encourage marginal

projects (Johnston & Johnston).

A CA is a long-term contract giving the Contractor sole right to operate on the

concession area in question whilst only paying Royalty and Taxes to the Government.

Royalty and tax rates are set at the region or country level and rarely vary by

concession. The CA system is viewed as skewed towards the Contractor (Løvås &

Osmundsen) for two reasons: it does not allow for Government participation in the

project, and the Contractor owns the mineral resources. Additionally, in a CA, the

Government can change the fiscal terms at any time as they are area-wide and not

included in the contract (McClean, 2023). This creates uncertainty for the Contractor

in projects that may span 40 years.

Contractual systems, of which PSAs are a modern example, date back to early

1800s France and are rooted in the idea that the Government should retain ownership

of mineral resources for the benefit of the citizenry (Kaiser, 2006). The PSA was first

used in the 1960s in Indonesia, at the time of the country’s independence, in response

to local hostility towards the foreign oil companies who held CAs (Bindemann, 1999)

and a desire by the Government for more control of, and benefit from, their own

resources (Hassan et al, 2022). The critics were satisfied with the new contract as the

Government retained control of the resources. Contractors were initially hesitant but

since then PSAs have spread to every continent. The PSA, being a negotiated contract,

offers more flexibility on a project-by-project basis than a CA which is governed by

the industry rules and regulations in force at the time (Hassan et al).

Key Goals of Production Sharing Agreements

The key goals of a PSA offered by a government include attracting skilled

foreign investors in the form of energy companies, facilitating skill and knowledge

transfer to host country workers and companies, providing revenue for the

Government and an acceptable rate of return for the Contractor (Dirani et al, 2021).

Many developing countries lack the technical or financial capability to extract

hydrocarbons and hence, cannot create value from them (Dirani), and PSAs provide a

solution in several ways (McClean). In fact, the PSA is the most prevalent form of

O&G leasing contract found in developing countries (Hassan et al). The Government

keeps ownership of the mineral resources and through the leasing agreement (the

PSA) appoints the Contractor to operate the project. The Government shifts the risk of

an unsuccessful project to the Contractor, who pays all the capital and operating costs.

In return, the Government allows the Contractor to recover those costs from a

successful project and share in the production and financial success (Sunley).

This research aims to investigate the utility of these points outside the context of

O&G by applying them to OWE.

National Oil Companies

Unlike any other hydrocarbon leasing contract, the PSA is frequently seen in

conjunction with the establishment of a national oil company (NOC) to participate in

the projects and receive the skill and knowledge transfer. Tracy et al (2011) discuss

several reasons why governments choose to set up a NOC. These include socio-

economic goals such as increasing employment, building infrastructure, and

improving technological capacity. These goals also include minimising government

bureaucracy - through its participation, the NOC monitors the private companies

managing projects, rather than needing a government department for oversight.

Hassan et al state that the joint committee of the international oil company (IOC) and

NOC are critically important in monitoring the operations of an oil or gas field. Also,

a NOC, through its participation as an equity holder and partner in a project, can gain

economic rent as a contractual party and increase the amount of cash that flows to the

Government above pure government receipts from Royalty, Profit Share, and Taxes

(Hassan et al). NOCs were prevalent throughout the developed world in the last

century – there were over a hundred created between 1920 and 2003 (McPherson,

2003). The early goals of the NOCs were nationalism, energy supply security (direct

control, not merely through legislation), job creation, manufacturing capacity and

infrastructure building, and policy implementation (Tracy et al, McPherson), and those

goals still hold for today’s NOCs.

It is true that NOCs frequently manage the downstream industry alone, but they

rarely attempt exploration or production (upstream) projects without an IOC partner

due to the sheer costs and risks (McPherson). The fact that developing countries (the

typical users of PSAs) have neither the needed expertise nor capital to run an upstream

O&G project (Hassan et al) makes the IOC-NOC partnership engendered by the PSA a

valuable one. Indeed, NOCs are expected to gain knowledge and expertise from the

IOCs through these partnerships, the governance of which is written into the PSA

contract (McPherson). The same is true for infrastructure building, job creation, local

manufacturing capacity improvements (all known as local content), training,

scholarships, and grants (Pongsiri). The Government also benefits by having a degree

of control over operations through the NOC and it may retain some mineral resources

for local use or sale (Majd, 2020).

Antolin et al’s case study (2013) contrasts Bolivia’s and Brazil’s NOCs and

concludes that a functional NOC (such as Petrobras, in Brazil) can improve the

country’s hydrocarbon industry performance by gaining better control of the country’s

natural resources compared to a country with no NOC. Brazil has also improved

innovation and the use of local suppliers in the industry through its contract terms

requiring those activities of foreign investors, in partnership with the NOC. The

authors also point to the fact that an NOC with strong technical, operational, and

financial capabilities will perform as a strong regulator and make that country

attractive to foreign investment. The converse is that a less capable NOC (such as

YPFB, in Bolivia) can lead to harsher regulations and fiscal terms and in turn,

discourage foreign investment (Antolin et al).

In 2003 NOCs controlled 90% of the world's oil reserves and 73% of the

production (McPherson) and are still widely seen in developing countries. Of the 82

countries with PSAs, 56 (68%) have NOCs which participate alongside IOCs in O&G

exploration, development, and/or production (Wood Mackenzie). From the same data

source, 51 (91%) of PSA-using countries with NOCs are developing countries.

Conversely, NOCs and PSAs are rarely seen in developed countries (only Brunei,

Thailand, Malaysia, Oman, and Trinidad & Tobago), most preferring CAs and

government oversight but not participation in the projects. The use of PSAs and

NOCs in the three aforementioned Southeast Asian countries dates back to the

proliferation of PSAs in the area after their invention by neighbouring Indonesia – at a

time when they were developing countries themselves. Since the early days, some

wealthy nations - the UK, France, Norway, Spain, Italy, Belgium, and others have

chosen to partially or fully privatise their NOCs, (Tracy et al) and no longer consider

the reasons for retaining a NOC important enough to bear the cost of running the

company.

This research intends to examine whether these advantages of NOCs within a

PSA framework could be applied to OWE.

Revenue and Profit Sharing

Government revenue in CAs comes in the form of bonuses (for access), royalty,

rentals (a minute portion of the total cash flow), and taxes. The PSA, on the other

hand, is designed to share the profit between the Government, who is the resource

owner, and the Contractor who takes on the expense and the risk of exploring for,

developing, and producing the hydrocarbons (Luo & Yan). Over decades PSAs have

been developed and refined and are considered the most functional method of sharing

wealth over a broad range of project characteristics (Johnston & Johnston) such as

hydrocarbon deposit size, cost to develop, hydrocarbon price, geologic risk, and

others. Costs and risks may be borne by both parties when the Government

participates in the projects through its NOC, or solely by the Contractor.

PSAs are designed to provide the resource owner income at various stages of the

project. This includes royalty from gross revenue, a share of the profits after costs

have been repaid to the Contractor, and taxes; that is corporate income tax plus, on

occasion, a special petroleum tax, on net income (Johnston & Johnston). The

calculation of the royalty, the fund from which costs can be recovered, the profit split,

and the special petroleum tax are often flexible and depend on prevailing hydrocarbon

market price, hydrocarbon resource size, project rate of return, or other factors. This

flexibility ensures that an economically marginal project can still be profitably

developed by the company. It also ensures that in the case of a hugely profitable

project, it is the resource owner and not the Contractor who receives most of the

excess profits above a certain rate of return to the company. In a CA, the resource

owner often financially benefits only from a royalty on gross revenue and corporate

taxes and does not share directly in the profits.

Rarely have countries switched between fiscal system types, but Brazil is one

example. The first contracts on offer to foreign companies, from 1995, were CAs of

the type used in developed countries such as the USA, UK, and Norway, and were

inspired by those systems (La Macchia, 2017). However, despite world-class

geology, Brazil was not able to attract foreign super-majors’ interest as it hoped. In

2010 the country switched to PSAs which were intended to be more flexible under

differing

project conditions. They also permitted the NOC, Petrobras, to participate alongside

the foreign companies in the projects and enhance its technical capabilities. Brazil

now has fifteen areas leased under PSAs (ANP, 2023).

One objective of this research is to study whether this revenue and profit-sharing

mechanism could have potential for OWE.

PSA Fiscal Terms

The four key terms in a PSA that affect the division of revenue are Royalty, Cost

Recovery, Profit Oil/Gas, and Tax (Kaiser & Pulispher, 2005a; Bindemann, 1999;

Hassan et al). There may be other minor terms, either legislated or negotiated, such as

signature and production bonuses, value-added or sales taxes, special petroleum taxes,

import/export duties, uplifts, and rentals. Cost Recovery, Profit Oil/Gas, and Tax are

common to almost every PSA. These three terms, together with Royalty, are found in

70% of global PSAs and have the largest effect on the division of revenue (Wood

Mackenzie).

Royalty is calculated as a percentage of the gross revenue. The rate may be a flat

percentage, or it may vary with the life of the project, production rate, cumulative

production, or other factors such as project return or oil quality. (Wood Mackenzie).

Royalty is popular with Governments as it provides up-front revenue as soon as a

project commences production, but it can be regressive, rendering marginal O&G

fields uneconomic as Royalty is paid before costs are covered (Tordo, 2007), hence its

absence in 30% of PSAs (Wood Mackenzie).

Cost Recovery allows the Contractor to be repaid for capital and operating costs

from the post-Royalty Revenue. A Cost Recovery Ceiling (CRC) of less than 100% is

often set so that costs can only be recovered from a portion of the post-Royalty

Revenue. This ceiling is above 50% in 90% of PSAs (Wood Mackenzie). A ceiling

much below 100% can cause a regime to be regressive but does allow the Government

earlier cash flow than from Profit Oil/Gas or Corporate Income Tax (McClean).

Unrecovered costs each year are carried forward.

The Profit Oil/Gas Split provides the next division of revenue between Contractor

and Government. The Revenue after Royalty and Cost Recovery is split between the

two parties. The Profit Split may be a fixed percentage or on a sliding scale with

production or other factors. It is popular to base Profit Split on an “R-factor”, which is

based on project return – 35 out of 81 (43%) countries employ it (Wood Mackenzie).

This allows good flexibility to support marginal projects but also permits the

Government to enjoy most of the upside from highly profitable projects. Profit Split is

a progressive term if structured well, encouraging marginal projects by directing most

of that profit to the Contractor; but providing a windfall to the Government in the case

of a highly successful project.

Income Tax is usually legislated and often applied at the same rates to all

industries and all companies in a country. Rates can change as tax legislation changes

unless grandfathered into the PSA contract. Tax rates may be flat or progressive, with

higher rates for higher levels of net income. Countries wishing to tax hydrocarbons at

a higher rate than other industries will enact a resource rent tax, or special petroleum

tax (SPT), payable in addition (Sunley). Countries that use PSAs rather than CAs

make use of SPTs less often – there is no need, as the Government shares in the profit

through the Profit Split. CA users must tax the hydrocarbon industry at higher rates

than other businesses to gain a larger share of the profit. This special tax revenue

comes at a later stage of the project than the profit-shared cash flows from a PSA

(Sunley). Taxes can also be mitigated by the Contractor by depreciation and

deductions, potentially limiting the flow to the Government. Globally, only 16% of

PSAs include SPT whereas 41% of CA regimes include it.

Other more minor financial terms commonly found in PSAs and worth

mentioning include rentals and bonuses. Rentals are small payments made annually

based on leased acreage, and bonuses are paid at the time of leasing but are typically

only found in highly prospective areas (Sunley). Energy companies are not willing to

pay multi-million-dollar bonuses for risky land grabs, nor where there is little

competition for the land (Sunley). One final term also seen in PSAs is Uplift (Sunley,

Tordo). In this researcher’s experience, it is often used in new areas – new

geographies, new markets, new geologies, and new technologies. Uplift is a multiplier

applied to costs before Cost Recovery to encourage the Contractor to make the

necessary investments. The Uplift allows the Contractor to recover costs more quickly

and gain from the time value of money. This term would not be seen in a highly

profitable or well-established area but is often included in contracts in geologically

risky or technologically expensive areas such as marginal deepwater fields in Angola

(Wood Mackenzie). Uplift has utility in this research as applied to OWE in

developing countries: a new technology in new geographies.

Non-fiscal terms to be found in PSAs which nonetheless affect the project cash

flow include local content obligations and domestic market obligations (DMO). In the

former, the Contractor must provide training for, or employ nationals, or use local

goods and services (Tordo, Tordo et al). In the latter, the Contractor must sell a

proportion of the production to the local market, often at a discounted price

(Bindemann).

Offshore Wind Terms

Subsidies and Incentives

Until the early 2020s, contract fiscal terms governing wind energy projects were

based on the Government offering subsidies and incentives to develop projects.

Examples are found in Europe (Higgins & Foley, 2014; Reichardt & Rogge, 2016;

Fristrup, 2003), Asia (Anwar & Mulyadi, 2011) and the USA (Anwar & Mulyadi,

Mormann, 2014). In Europe and Asia, which have producing wind farms, these terms

are collectively known as feed-in tariffs (FIT) which are part of a policy that mandates

energy utilities to pay a premium to producers that generate electricity from renewable

sources (Mendonça, 2012). They are found in two forms: feed-in premia and feed-in

prices (Alolo et al, 2020). A feed-in premium in offshore wind is a government-

provided premium above the market sales price of power whereas a feed-in price is a

government-guaranteed sales price for power, also known as a contract for differences

(CfD). Burer & Wustenhagen (2009) found in a survey of sixty investment

professionals that feed-in prices are the preferred subsidy method, and Polzin et al

(2019) confirmed this in a review of ninety-six studies on the impact of policy on key

investment metrics. However, the efficacy of feed-in premia is debated, with Alolo et

al (2020) finding it ineffective based on discounted cash flow modelling in 27 EU

countries whereas Reichardt & Rogge (2015) had previously found it to be effective

Germany. Bunn & Yusupov (2015) assert that feed-in premia will become ineffective

as wind power supply exceeds demand.

Community Benefits and Compensation to Local Residents

In several countries, provisions are made to compensate local communities for

the burdens felt living with a wind farm, but examples for OWFs are rare (Rudolph et

al, 2015). Most are related to onshore wind farms which have been extant for longer

and are more likely to inconvenience local communities. In Denmark, two schemes

are underway applying to both onshore and offshore wind; a third was halted after ten

years in 2018. The first is a property-value loss scheme and the second is a co-

ownership scheme (Jorgensen et al, 2020). The authors describe the former as

financial compensation for diminished property value due to being located near a

wind farm, and the latter as financial returns through holding an equity share in the

business. The latter can be offered to a wider community than the former which is

constrained to neighbouring properties. The authors find that these purely financial

schemes have limited success unless alongside recognition, another key dimension of

environmental fairness.

Financial assistance was also the most common benefit encountered by Rudolph

et al in the UK and the Netherlands. The UK is considered a leader in community

benefits from onshore wind in that GBP5,000 must be spent per MW of installed

capacity (RenewableUK, 2013). However, offshore, the amounts were often less

generous, and participation was considered voluntary (Glasson, 2020). As well as

funds which provide financial support to the community, schemes which directly

provide apprenticeships and education were seen in the UK. Ownership in the facility

was seen by the same authors in Germany and the Netherlands.

In 2014 (Rudolph et al) the first community benefit agreements in the USA were

seen during leasing off Massachusetts, but these did not specify up-front amounts or

activities which would aid the local population. The bids in California’s first offshore

wind lease round in 2022 (Ayoub, 2023) also included community benefit agreements,

this time with more specific terms, supporting initiatives for workforce training and

supply chain development. This last example, having specific development goals and

not just financial support, is akin to some of the terms seen in PSAs in O&G in

developing countries.

The traditional PSA, as described previously, mandates specific benefits that

must be built for the host country such as infrastructure. It also obliges funds, such as

for training and education, which are financial benefits. A PSA, being an enduring

contract, would provide these benefits to the communities for the life of the wind

farm. The communities would be protected from a change in legislation that might

cause a loss of those advantages or compensation, the same way PSAs protect those

rights attached to O&G projects.

There are no national energy companies to mirror the NOCs in the O&G industry,

though two developed country NOCs, Equinor and Ørsted, have diversified into OWE.

Both companies are operators and are not in the position of needing to learn from

experienced operators through a PSA-type arrangement. There are no instances of

developing country governments directly benefitting from OWE in the manner that

they do through PSAs in the hydrocarbon sector.

Existing and Mooted Leasing Mechanisms

Laido et al (2022) make the important point that providing developers of OWFs

access to the seabed is a crucial piece of the puzzle. Gonzalez et al (2020) performed

case studies on the offshore wind leasing mechanism in Denmark, the UK, and

Germany. From 2013 to 2019 those countries were the top offshore wind power

generators in Europe; even in 2022, they are in the top four, joined by the Netherlands

(IRENA, 2023). All were characterised by a mixture of tenders (bidding rounds) and

direct negotiation, by multi-phase licences corresponding to different activities, and

historically, by FIT or CfD incentives. Government participation in the projects was

not mentioned, indeed, none of these countries’ governments participates in O&G

projects. Last century, Denmark and the UK had state O&G companies, now known

by the names Ørsted and BP. Ørsted is no longer active in O&G and BP has been

privatised. Interestingly, both are significant players in offshore wind today.

Community benefits as discussed in the previous section do feature in the leasing

practices of all three of these countries.

In China the approach has been uncoordinated, involving numerous ministries

and agencies (deCastro et al, 2019) making the meteoric rise in installed capacity

even more remarkable. Hybrid systems of approvals exist, involving municipal,

regional, and central energy departments, and long-range plans are made at the county

and prefecture levels (Hughes et al, 2022). As such, this researcher has not been able

to elucidate a clear picture of the leasing mechanism(s) in use. As outlined later in the

section “Beyond Subsidies,” FITs were in existence until late 2021 but discussion of

the type of lease – whether or not it mirrors an O&G CA-type agreement - does not

appear to be extant.

The USA has been slow to adopt OWE (Firestone et al, 2015). There was a push

in 2005 with the US Energy Policy Act which required the Department of the Interior

to devise a leasing system for wind energy in offshore waters and they had a leasing

mechanism in place by 2009 (Firestone et al). By late 2021 (Comay & Clarke, 2021)

only one commercial-scale (though small, at only five turbines) OWF was operational

at Block Island, offshore Rhode Island. It is in state waters, not federal. This is despite

leasing having taken place in federal waters annually from 2012 to 2023 except in

2020 and 2021 offshore Delaware, Maryland, Massachusetts, New Jersey, New York,

North Carolina, Virginia, and California (BOEM, 2024). The first Gulf of Mexico

OWFs were approved in 2024 (Budin, 2024) in Louisiana’s state waters; the state

recognised that although their economy has been heavily dependent on O&G for

decades, renewables will play a large part in their energy supply future. These USA

lease auctions are modelled closely on the offshore O&G lease sales which have been

ongoing since 1954. Firestone et al attribute this slow start for OWE to three factors.

First, the burdensome offshore O&G regulatory regime being inappropriately applied

to offshore wind – it is onerous in order to protect against the potential catastrophic

disasters that can beset O&G installations – which do not apply to wind farms. A fit-

for-purpose regulatory regime would be more desirable. Second, the leases place an

emphasis on up-front lease bonuses committed at the auction, which can have a

regressive effect on the project's economics. A preferable system would be to extract

royalties from the revenue of the operational OWF, though the government would

wait

longer for their portion of the revenue. Third, although the USA had an interest in

OWE, it did not put strong policies in place at the time to encourage offshore wind

over hydrocarbon energy sources.

The USA has chosen not to offer price-supporting tariffs like China and Europe

have. Instead (Energy.gov, 2023), tax credits are available to both onshore and

offshore wind farms. Three types of credit reduce corporate income tax burden; based

on generation from the OWF, capital expenditures made, and power capacity built.

This method has the least regressive effect on project economics but is coupled with

highly regressive up-front lease bonuses.

A notable fact about all the leases offshore USA is this payment of large up-front

cash sums. The five leases in the 2022 California bid round went for US$757MM

(Ayoub) and the sale offshore New York and New Jersey in the same year attracted

bids of US$4.37 billion (Volcovici, 2022). The author points out that one advantage to

the up-front sums is that the money goes into the Treasury and benefits taxpayers.

However, marginal projects can be damaged by high up-front payments, and in the

future less-promising areas may not receive any bids. A contract such as a PSA, where

less revenue flows to the government initially, encourages marginal projects. If the

PSA is structured correctly, the government can more than make up for forgoing early

cash by taking the largest portion of the upside later in a successful project (Mohamed

et al, 2022a). Naturally, no one gains revenue from a project that does not go ahead

because the contract terms inhibit it economically.

With respect to offshore wind California could be compared to a developing

economy for the following reasons. It has no local supply chain like the East Coast of

the USA, there are no power purchase agreements (PPA) in place at the time of

bidding to ensure sales price, the permitting process is complex, and there is a lack of

grid infrastructure (Ayoub). Add those factors to the need for floating OWFs which

are more expensive, risky, and based on much newer technology than fixed-bottom

farms and it is a wonder there are any interested developers. There were bids,

however, albeit lower than East Coast bids. This illustrates that even in new and

difficult areas, there is a desire to build OWFs when the conditions are promising.

The USA’s government has no desire to get involved in the building or operating of

offshore wind projects – nor do they in O&G projects. The only area in which they

may choose to involve themselves is upgrading the grid (Comay et al), as it is

typically multi-state infrastructure. A leasing model which contained a NOC (or

NEC) would therefore not suit the USA.

Australia and New Zealand are even newer to OWE than the USA. Australia has

leased only one area to date (Coull et al, 2022). The leasing regime comprises three

licence types: exploratory, feasibility, and commercial. The lease in question was

granted in 2019 as an exploratory licence, and the developers hope to move it to the

feasibility phase in 2023. The plan is to begin construction in the mid-2020s for first

power delivery in 2030 (Star of the South, 2023). The up-front payment for these

licences has not been published - and the licence was designated rather than subject to

a bidding process. The wind farm proposed on the acreage will have a capacity of over

2 GW and involve 250 turbines, over double the capacity of the average farm starting

up in 2020-2022. Rapid increases in capacity and technology should be expected in a

maturing industry. Australia recognises the positives of offshore wind in terms of

reliability (relative to onshore wind and solar) and the fact it will be low-cost by 2030.

It sees as an advantage the ability to build OWFs near demand centres (coastal cities),

and the possibility to transition O&G workers to the new energy installations. Rispler

et al (2022) found that at the lower end of cost estimates given for future installations,

OWE would not only displace hydrocarbons but also onshore wind and solar. The

authors suggest that although net zero is a goal for 2050, due to the difficulties in

replacing some uses of energy with clean sources, in Australia, heating and electricity

must be net zero by 2035 to 2037. This leasing regime, like the USA and Europe, does

not offer the Government a chance to participate in the project and offers revenue only

at the bidding stage, and in taxation of the Contractor as a going concern. No mention

was made of community benefits. Perhaps no coincidence that Australia also uses CAs

for O&G leasing (Wood Mackenzie).

New Zealand realises it has great wind potential and considerable coastline per

capita. The country is currently designing its regime for OWE (Coull et al) which will

partly be based on best practices from the O&G industry - but with changes where

merited (unlike the USA). The country is partly looking to Australia’s regime for

ideas. New Zealand is cognisant of the issues faced by offshore wind and is ready to

address them (Venture Taranaki, 2020). It realises transition is needed, which involves

electrification, and as the population and the economy grow, power demand will

double in 30 years. The intermittency can be offset with hydro generation which is

already established in the country. The need for grid upgrade is acknowledged in all

but the most modest OWE scenario. New Zealand is looking to the future with

innovation – realising it has wind potential that could outstrip domestic demand,

Venture Taranaki outlines the idea of power storage via hydrogen. Although the initial

wind farms in the study are modest, at 200 MW to 800 MW, the two areas in the study

could deliver 2.4 GW to 12 GW if fully developed - this higher number matches the

UK’s current operational capacity. Like Australia, no mention is made of community

benefits or an option for the Government to participate – and New Zealand also uses a

CA regime for its O&G leasing.

In 2020, Gonzalez et al contributed to the design of the offshore wind leasing

regime in Brazil with their paper combining the current practices for offshore wind

leasing in Europe with the various regimes in use in the Brazilian O&G sector. Their

suggestion was akin to a CA. Indeed, in 2022, the regulatory framework was passed

into law (Palmigiani, 2022) and uses an O&G-like CA with a signature bonus payable

up front, a Royalty on gross revenue, and potentially corporate income taxes if the

company is financially successful. According to Palmigiani, the bonus and Royalty

system was criticised as it may hinder investments, and the Royalty rate had to be

reduced before the regime passed into law. There is no evidence that a more

progressive PSA-like system with the sharing of profit after costs was considered.

There is also no mention of non-financial benefits such as seen in PSAs.

Other countries preparing for their first leasing for OWE include Norway,

Lithuania, Uruguay, Portugal, Colombia, Estonia, and Ireland (Emanuel, 2023), and

promisingly, one of those, Colombia, is a developing nation.

This research aspires to provide an alternative to the O&G CA-type leasing

mechanisms seen above; one that may be of special interest to developing countries,

the primary users of PSAs. It will quantitatively investigate applying O&G PSA

contract terms to an OWF and through qualitative enquiry will uncover the most

important provisions of the PSA.

Power Purchase Agreements

A PPA governs the relationship between the seller of power (the generator) and

the buyer (the utility company), and covers construction, operations, maintenance,

interconnection to the grid, and any third-party involvement in the project (Bruck et

al, 2018). These factors contribute to the cost of power generation which must be

covered (plus a rate of return) by the buyer (Bruck et al). PPAs are on the rise in

conjunction with offshore wind globally and commonly encountered in the USA,

Europe, and Latin America (Bruck et al). They provide more certainty to the project

economics by setting a fixed sales price. Though as Bruck et al point out, uncertainty

still exists for the seller in two scenarios. First, if their OWF underperforms they may

have to purchase power on the spot market to fulfil commitments. Second, if the buyer

exercises their right to cap their power offtake, and the seller cannot sell the excess

power, it drives up the seller’s per unit costs.

PPAs are often seen in combination with renewable portfolio standards (RPS) in

which a government sets a percentage of power nationally that must be obtained from

renewable energy sources. The PPA allows the buyer to cap the amount of electricity

purchased from the renewable sources in order to limit exposure to often higher prices.

Hydrocarbon and nuclear power remain cheaper in some situations, and certainly were

historically.

The PPA is downstream of the generator’s leasing of the land or seabed on which

to site the wind farm. A generator could lease from a government under a PSA

arrangement which divides the revenue between the resource owner (Government) and

the operator of the OWF (Contractor). The Contractor could then enter a PPA with a

buyer to increase the sales price certainty. The Government, through its NEC, could

still participate in the project and increase its technical capabilities. Hence, a PPA

could be used in conjunction with PSA-like terms for an OWF. This research assumes

a variable price for power; the addition of a PPA would simply remove some of that

price uncertainty and could be the subject of future research.

Beyond Subsidies

Oil and Gas Industry

Incentives such as tax credits have indeed been available to the hydrocarbon

industry since its inception. Countries such as Norway offered incentives in the early

years then phased them out as the industry became economic without assistance

(Ryggvik, 2015). At times, incentives have been phased back in to counteract

prevailing economic conditions such as the oil price crash in the 1980s, e.g., in

Norway (Lund, 2014) and the UK (Band, 1991). In some countries such as Indonesia,

hydrocarbon incentives are prevalent still today, however, it has been shown that these

do not achieve the goals they were intended to (economic growth, poverty reduction,

and energy security), and moreover, decrease the government’s revenue (Pradiptyo et

al, 2015). Metcalf (2018) demonstrated that in the USA, removing hydrocarbon

incentives would increase government revenue, though it would not mitigate energy

demand, which in turn would help the environment. The global O&G industry still

enjoys some subsidies in the form of uplifts for recovering or depreciating capital. At

the same time, it finds itself taxed more harshly than other industries in 49 countries

through SPT, including twelve countries whose SPT rate is 50% or greater (Wood

Mackenzie, 2022). The same source reveals that seven of those countries enact the

highest tax rate when price or revenue are above a certain level, but The Netherlands,

Denmark, and Norway have flat rates of SPT at 50%, 52%, and 56% respectively.

Renewable Energy Industry

Babcock (2013) concluded that subsidies can safely be removed in the USA when

the renewable fuel source can compete economically with hydrocarbon fuel sources.

Their study was into biofuel ethanol - the renewable energy source with the longest

history – and the first to consider removing subsidies. Burer & Wustenhagen heard in

2009 from their survey respondents that in general, investors expect renewable energy

subsidies to be removed in the future and that project economic base-cases are

calculated without subsidies as the timing of removal is unknown. Xiang et al (2021)

point to a second reason for the removal of subsidies – in China, renewable energy

increased thirty-fold from 2010 to 2018 and the government can simply no longer

afford the subsidies on this higher activity base.

As renewable energy sources become economic without subsidies, government

revenue could benefit from halting the incentive schemes (Xiang et al). This will be

an important milestone, especially in less well-off countries that wish to implement

renewable energy projects but cannot afford to subsidise the industry. As renewable

energy sources become more widespread, technological advances occur, and the costs

of installing capacity fall, subsidies and incentives may no longer be necessary

(Johnston et al, 2020, Lee & Xydis, 2022).

The post-subsidy age will benefit developing countries such as Vietnam (Zimmer

et al), which in 2014 was only able to offer a 1c/kWh subsidy on power from

renewables, barely improving their competitiveness compared to domestic

hydrocarbon supply.

Falling Costs of Offshore Wind

The World Bank (2019b) found that between 2000 and 2019, offshore wind

frequently cost US$60-200/MWh, three times that of onshore wind energy, but the

trends are downward (Figure 3). Eicke et al’s calculations confirmed that.

Figure 3: Levelised offshore wind tariffs to 2030

By 2018 (World Bank 2019b) China was consistently breaking the US$100/MWh

barrier for installed capacity and in Europe, auctions were revealing a cost below that

same barrier for capacity to be installed shortly thereafter. The IEA found (IEA, 2020)

that in 2019 onshore wind was the cheapest method of generating electricity from any

fuel source, and whilst OWE is more expensive, it is dropping faster as it is a newer

technology. IRENA (2019) points to several factors that will continue to drive costs

down – economies of scale as offshore wind penetrates the market, more competitive

supply chains, and technological improvements. They also observed that onshore

wind is already cheaper than new hydrocarbon-fuelled installations in many areas. In

their study about the removal of subsidies, Peng & Wu (2022) found that OWE cost

only two times as much as onshore in China in 2020. This agrees with the IEA 2020

view that onshore wind costs US$40-65/MWh and offshore US$75-110/MWh.

Auctions for capacity to generate OWE in 2023 in the UK now have a cost below

US$50/MWh (IEA, 2020), and in Germany and the Netherlands, zero-subsidy projects

have been bid (IRENA, 2019). Mathews et al find that the cost of even OWE is now

falling below hydrocarbon sources such as coal and gas, eliminating prior non-

competitiveness. NREL (2019) points to the 68% decrease in the cost of fixed bottom

OWE between 2010 and 2019 and the projected further decline of 42% by 2030 (from

US$89/MWh to US$51/MWh). Many countries are gaining an interest in offshore

wind due to the decrease in cost. For example, in the last decade Australia did not

consider OWE viable in its renewable energy mix but now finds that it will be cost-

competitive. Hence, they offered the above-mentioned lease in 2019 with plans for

more, producing up to 40 GW by 2037 (Rispler et al).

Mathews et al also see floating OWE, currently more expensive than fixed bottom

turbines, falling below US$60/MWh by 2030, and Casey cites areas offshore USA

costing US$50-75/MWh in 2023. Casey reports on the US Department of Energy’s

belief that due to technological advances, and she describes three, even floating

offshore wind can reach US$45/MWh by 2035.

Fixed bottom turbines can operate at water depths of up to 50m, often near the

shore. Some countries’ shoreline falls off more steeply and floating wind will be the

best option. This study will only consider fixed bottom turbines, but a future direction

would be to include floating wind farms.

There has been some discouraging news from the USA in 2023 about two

planned New Jersey OWFs being abandoned due to high inflation, rising interest

rates, and supply chain bottlenecks (Ørsted, 2023). However, no transition to a new

technology is ever smooth, and this researcher is going to continue to monitor the

situation and assume that one setback does not condemn the whole industry.

Rarely quantified, but nonetheless important, are the cost savings associated with

a shift to renewable energy (IRENA, 2019). The expenditure associated with the

health impacts of poor air quality and the clean-up and rebuilding after climate events

could be saved, decreasing the overall cost of power from renewable energy sources

versus hydrocarbons.

This cost decrease, and the associated end to the need for subsidies, will make

OWE available to countries who can neither afford to subsidise, nor develop sources

that cost more than the existing energy supply.

Offshore Wind Energy Subsidies

Denmark phased out subsidies paid for by the taxpayer (FITs) in 2000 (Fristrup)

though it appeared the move was premature as offshore wind generation capacity

installation stalled, and a new FIT-like instrument was introduced in 2008 (WWEA,

2018a) – and capacity installation resumed (IRENA, 2013a). Then in 2021, for the

first time, a Danish auction of seabed for an OWF resulted in cash inflow to the

government rather than the promise of cash flow from the Government to the

Contractor (Laido).

Similarly, the Netherlands introduced FITs in 2003 to then decrease them in 2005

when the government realised it could not afford them (WWEA, 2018b). The first

acreage to be leased without a subsidy was in 2018 (Laido) and the authors’

expectation at the time was that it would be the first operational OWF to involve a

payment to the Government rather than from it. Indeed, the winning bid by Oranje

Wind Power II C.V. included a one-off EUR50MM payment, and the farm will

receive no subsidies (Jetten, 2022). The farm, Hollandse Kust Zuid, was inaugurated

in late 2023 and is expected to commence operations in 2024 (Power Technology,

2023). Power sales are subject to a fifteen-year PPA.

Spain (WWEA, 2018c) had the same policy, offering FITs until 2013, at which

time the country switched to a tax on wind power, and new generation capacity

slowed. Despite this, Spain was still the fifth-largest combined onshore and offshore

wind producer in 2022 (IRENA, 2023). Conversely, Germany has been remarkably

successful – in 2021 the country was the third largest combined onshore and offshore

wind producer behind more populous China and the US. The government offered

stable, long-lived, and transparent policy and terms for several decades contributing to

the success (WWEA, 2018d). Although FITs halted in favour of auctions, the industry

survived, though small companies are likely to be driven out by this change (WWEA,

2018d). Jamasb & Sen (2022) and PWC (2018) point to zero-subsidy bids in Germany

as far back as 2017.

The UK has tried four policies since the 1990s – lacking stability and longevity –

but in 2021 was still sixth in the world in combined onshore and offshore generation,

and second in offshore alone (IRENA, 2023). From 2018 subsidies were phased out

and the WWEA predicted a fall in new licensing once the existing licences are

brought into production and older facilities are taken off-stream (WWEA, 2018e).

However, this trend was not evident judging by Westwood’s report that the 2022

Round 4 auction was the largest ever, or by late 2023 with the Government’s

announcement to offer sites for over 15 GW of floating wind in the Celtic Sea (The

Crown Estate, 2023). The 2021 seabed allocation auction ushered in a new era in the

UK with bids made including payments to the Government of EUR 250,000-

800,000 /km2 (Laido) during the wind farm development phase.

Groom (2023) reports that the large bonus amounts paid to the USA Government

in 2022 for offshore wind leases indicate the faith energy companies have in OWFs

which do not enjoy price support or FITs, though as previously discussed they do

qualify for production and investment tax credits.

In China, central government subsidies were removed at the end of 2021, and

Mathews et al opine that the policy change is an indication that the Government is

confident in the ability of the industry to be cost-competitive and achieve growth and

stability without assistance. Three provincial governments are not so confident and

have rolled out local subsidies which apply to wind farms coming online between

2022 to 2024 (Chen, 2022). These subsidies have less value the later the farm comes

online and will only run for the first 10 years of the farm’s life. China’s offshore wind

power targets have 2025 as the stated endpoint, so perhaps it is not a coincidence that

the subsidies apply to farms coming online by the end of the previous year. Chen

states that it is not the subsidies that are the main driver for OWE growth in China,

though the new construction activity has dropped to a more normally expected level

after the cessation of the national subsidy.

The Future

Subsidies will soon be behind us as wind farms under subsidy schemes make up

an ever-smaller proportion of total OWE. IRENA (2019) expects that offshore wind

will be subsidy-free by 2030 in many global markets. Jansen et al (2020) expect that

Europe will be subsidy-free by 2023 and judging by the earlier findings in the UK and

the Netherlands this goal has been achieved between 2021 and 2024. Kitzing (2023)

talks of the changes in approach required when offshore wind projects switch from

needing financial support to becoming net contributors to state finances. In her case

study of Denmark, she points to the policy shift recognising the need for subsides to

one of wanting to maximise income and profit for Danish society. Indeed, Bloomberg

Energy Voice (2022) reports that power from Triton Knoll, off the UK, is so much

cheaper than power from gas-fired plants that it is paying the excess between the cost

of generation and the sales price to the government. The article also points to Danish

Hornsea 2, which will soon be able to undercut Triton Knoll’s price by 22%. This

illustrates the shift from an industry requiring subsidisation to one that pays a negative

subsidy – or tax. An initial step in this regard may be a two-sided CfD (Jamasb & Sen)

which will attract a subsidy when the power sales price is below the cost to produce

but pay a tax when the price exceeds the cost.

However, CfDs are an adjustment to the existing contract type rather than a

radical change into a new era. This existing contract type has similarities to the

regressive fiscal regimes seen in O&G CAs where fees are fixed regardless of the

profitability of the project. Examples are auction bids (or lease bonuses), rentals (or

acreage-based fees), or rates applied to gross revenue (such as royalties or annual

per MW fees. These taxes on gross revenues can render a marginally economic

project sub-economic and halt that project. As Laido et al point out, this up-front

loading of costs can affect financing, competition, and commitment to projects. A

structure where the Government takes its cut from the profit rather than the revenue

would encourage marginal projects.

Any wind farms built with contract terms under discussion in this research would

not be operational until the mid-2030s, therefore zero-subsidy leasing is a good

assumption.

The Opportunity to Structure New Contract Terms

As Jamasb & Sen point out, there is an opportunity to consider the distribution of

revenues from OWE in the post-subsidy world. Perhaps an entirely different contract

type could offer an interesting way forward. Then it may be prudent to look to the

industry that has been producing energy offshore for decades – namely the O&G

industry. Snyder & Kaiser (2009) state that the regulatory system most applicable to

OWE is that of O&G due to the similarities between the industries – expensive

offshore installations that provide energy from offshore resources. As Stephanie

McLellan said of the overlap between offshore O&G and offshore wind, “It’s the same

industry, just using a different fuel” (Magill, 2019, paragraph 13). Why not also look

to the leasing system for best practices that could be applied to offshore wind?

Markwell et al (2014) posit that PSAs may be seen in the future in other (non-oil or

natural gas) types of hydrocarbons such as methane hydrates or coal bed methane, so

why not other forms of energy too? The argument may be made that oil and wind are

too different. Oil and gas can only be exploited once whereas wind is renewable – but

that fact need not drive the leasing mechanism. Oil is a transportable commodity

whereas wind-generated power must be used in situ or close by. But gas, which is

harder to load onto a carrier vessel and export, is also produced offshore and

transported by pipeline for use in regional markets. Additionally, wind-to-hydrogen as

a means to store and transport wind power is an emerging technology (He et al, 2022).

The hydrogen can be used at the installation or piped to shore in either former gas

pipelines or new-build pipelines. In this way, OWE could one day function like the

liquified natural gas (LNG) industry which converts the energy source (gas) to a liquid

to enable transportation to areas of demand. Spezakis & Xydis (2022) see offshore

wind-to-hydrogen as an economic method to transport power from the USA’s

Western Gulf of Mexico region to coastal populations as there are many unused gas

pipelines in the area. The authors tell us that since OWFs may be far offshore, the

cost of new submarine cabling will be greater than re-purposing gas pipelines.

Further, as hydrocarbon production decreases and more O&G infrastructure is

abandoned, more pipeline infrastructure becomes available. The differences in storage

and transportability of oil, gas, and wind do not dictate the leasing mechanism.

A second argument for the differences between OWE and O&G is that

hydrocarbons often provide a large portion of foreign income for a government

(Sunley) whereas power, if used locally, will not. Again, this need not drive the

leasing mechanism. The differences in the energy types do not constitute a reason not

to study the transferability of PSA best practices to OWE.

It is true that mathematically speaking, the same financial outcome can be

achieved by using a CA or a PSA – depending on how the terms and rates are

designed (Sunley) – so a hydrocarbon-style CA could equally be applied to offshore

wind. The result would be similar to existing offshore wind leasing in developed

countries. However, the CA may replicate the PSA for one specific project – but if the

project’s parameters change, the outcome may be very different. The CA, being

regressive, will harm marginal projects and dissatisfy Governments in highly

profitable cases. Additionally, Governments would be missing out on the non-

financial benefits detailed elsewhere in this paper that can be enjoyed when using

PSAs.

Olleik et al (2021) studied an innovative suggestion for diverting a portion of the

profit oil or gas in a hydrocarbon PSA to investments in renewable energy in

Lebanon. This provides a linkage between O&G company strategies to invest in

renewables and a developing country’s need to fund such projects. This was the first

suggestion in the literature that PSAs could be connected to the funding of renewable

energy projects but does not apply the PSA to leasing for an OWF. Hassan et al talk

about future PSA evolution as environmental policy develops, but like Olleik et al, do

not go as far as suggesting a PSA could be applied to more climate-friendly energy

sources.

Going further than those studies, and to the best of this researcher’s knowledge,

this current study is the first to explore the use of PSA-like terms for an entire

renewable project without a hydrocarbon element.

It should be noted that the use of PSAs does not preclude the current use of

auctions as the method to allocate seabed; the PSA forms the lease contract. Jansen et

al (2022) expect by 2030 that 97% of all seabed sites awarded will be by auction. The

PSA would govern the terms of the contract area for the lifetime of the project. Those

terms could either be directly negotiated with the Government or could be biddable in

an auction to foster competition. Both approaches are seen in the hydrocarbon

industry – 34 countries use PSAs in conjunction with licence rounds (bids), 21 directly

negotiate, and 26 use both (Wood Mackenzie). As discussed before, the use of PSAs

does not preclude the use of PPAs for the offtake and sale of power. In Germany and

the Netherlands, both seabed access and offtake are bid in the same auctions whereas

in the UK and USA, they are handled separately (Bloomberg NEF, 2022). Either way,

using a PPA to bring certainty to the revenue achieved from the project would work in

conjunction with a PSA governing the relationship between Contractor and

Government. A PSA-like contract, being the leasing mechanism, could also dovetail

with RPS or other renewable targets that a government might set – the targets set the

ambition, and the leasing contract enables its fulfilment.

Fair Sharing of Profit

Given the rapid drop in the cost to produce power from offshore wind, long-

term fixed-price contracts seem biased toward the Contractor. As profitability

increases, even over the life of a given installation, the resource owner or

Government may be

forgoing revenue from its natural resource (Snyder & Kaiser). The authors discuss two

forms of lease fees which are payable from gross revenue (royalties) or paid

regardless of revenue status (rentals) and their regressive nature. If these lease fees are

set at a low level to encourage investment, the resource owner misses out on the

upside in the case of a highly profitable project. PSA-like contract terms, which share

revenue at multiple stages of cash flow, can be structured to be highly progressive and

avoid these pitfalls. The Royalty, calculated on gross revenue, ensures that the

Government receives at least a minimum payment, and the Profit Split ensures that the

upside is enjoyed by both Contractor and Government (Sunley). Corporate Income

Taxes may not provide the same degree of profit-sharing for the Government if the

Contractor company has other losses to offset successful wind projects.

Fair Recovery of Cost

With its cost recovery mechanism, the PSA provides better assurance to the

Contractor that its investment will be recouped. Under a CA-like system, no allocation

is made from revenue to repay the Contractor for capital and operating expenditures.

Under PSA-like terms, the Contractor recovers costs before the remaining revenue is

split according to the Profit Split ratio. Recovery of cost is an attractive feature for an

investor willing to spend billions of dollars building an oil field or wind farm. Cost

recovery is clearly not a necessary condition, given that OWFs are being built without

this assurance, but this research aims to uncover whether PSA-like terms might be

superior to a CA-like leasing system in offering this feature.

An area of contention in O&G leasing is who pays for site identification

performed prior to an auction or bid round. This could involve industry-sponsored or

proprietary company seismic studies over an area of seabed to ascertain if it is

attractive for leasing. Under the CA system, there is no mechanism to recover those

costs, at best they can be considered tax deductions. Under a PSA, the expenditure

made before the contract was signed on the contract area can be considered allowable

for cost recovery under the PSA when eventually signed (Nichols, 2010). Indeed, they

are often recovered before any current operating or capital expenditure. Do et al

discuss this issue in reference to OWE, positing that the industry wants to identify

suitable sites for OWFs – in fact they prefer to undertake the surveys than let the

Government carry them out – but they do desire a mechanism to later recover those

costs. As in O&G, this would be more easily accomplished via contracts, like a PSA,

than through concessionary systems.

Encouragement of Marginal Projects

If lease fees are high, marginal projects will not be developed and both

Contractor and Government get no revenue - and the local area will have to source

power by another route. A more flexible contract type, like the PSA, which varies the

rates of revenue-sharing based on the profitability of the project, may be appropriate

for OWE. Its progressive nature would provide the opportunity for marginal projects

to be developed and for the Government to take large revenues from highly successful

projects. This has been recognised by Firestone et al in their analysis of the USA’s

offshore wind leasing regime though in the eight years since the authors made the

observation, the USA’s regime has not changed.

Community Benefits

Community benefits in existing offshore wind leasing agreements in Europe and

the USA were analysed previously in this paper. A PSA-like contract, including

O&G- like provisions for local development, would take these community benefits

further, sealing them into the contract for the life of the wind farm. Benefits could be

specified, unlike the financial amounts seen in the UK with no explicit project

attached. Developing countries could negotiate terms for education, training,

infrastructure building, and local content (jobs and use of local manufacturing

capacity). Infrastructure could be defined as industry- and supply chain-related, such

as port improvements; schools or universities for education and training; and civil

projects such as roads. Given the smaller profit margins of power generation projects

compared to successful O&G projects, it is likely that the scope of community

benefits per contract will be smaller than for O&G, but nonetheless desirable to the

Government offering the contract.

Knowledge and Expertise Transfer

Developing countries with natural resources, but which lack the capital and

human resources to develop those assets, favour PSAs (Markwell) as a way to inject

finance and expertise into the projects. This expertise transfer comes through the

establishment of the NOC and its participation in the projects. For OWE, a NEC could

be set up and participate in the PSAs with the international Contractors. Once there

has been an expertise transfer, the domestic industry can develop smaller projects that

would not attract the interest of larger international energy companies (Markwell).

The domestic market can be satisfied without reliance on outsiders. Almost a decade

ago,

IRENA (2013b) found that a market driver for wind energy, and a motivator for

countries to encourage the industry, is creating local economic value through jobs and

manufacturing. In 2019, the same organisation pointed out that the right education and

training are crucial to the establishment of local supply chains to facilitate offshore

wind projects. Similarly, they state that policy must encourage the development of a

workforce. The use of NECs and local content clauses in PSAs could support this

goal. In fact, this behaviour can be seen in the USA, for example at Vineyard Wind in

Massachusetts, where the developer committed US$15 million of funding for

education programmes and apprenticeships to build a wind energy workforce (Tyler et

al, 2022). This was an important factor in gaining local acceptance of the planned

installation. This kind of local benefit is commonplace in a PSA and would not have

to be tackled as a separate initiative. Study participants from the local community

around the proposed Kitty Hawk Wind Lease (offshore North Carolina, USA)

expressed an interest in discounted electricity for residents as a way to increase

acceptance of the development of a wind farm (Tyler et al). This is similar to the

DMO often seen in PSAs, where hydrocarbons are sold to the host country at a

discount to market prices.

Risk Sharing

It is true that the choice of fiscal regime in the hydrocarbon industry is driven in

part by risk and reward sharing – there are immense risks and costs in O&G

exploration, which concerns IOCs (Kaiser & Pulsipher, 2005a, Tordo et al, 2010,

Pongsiri). A company can spend hundreds of millions of dollars on a leased area: on

leasing, seismic studies, and drilling, and there is only a 10-30% chance this expense

will ever be recovered (Brez-Rouzaut, 2022). The cost of exploring for oil or gas – and

finding one drop of hydrocarbon - not for developing the field or producing it – can

run from US$200MM (million) to US$500MM for a 100 MM barrel field (Brez-

Rouzaut). By contrast, a US$4MM to US$10MM feasibility study can determine the

wind potential which drives the design and cost of an offshore US wind farm

(Finmodelslab, 2023 and Appendix 1) Despite this mismatch in risk, the PSA, with its

multiple revenue streams to the government and non-financial benefits including

local development and equity participation, offers an interesting alternative to the

leasing mechanism seen in developed countries. Further, as discussed earlier, with

the use of long-term PPAs, revenue can be made less uncertain.

This research is the first to explore the reasons why governments of developing

countries with untapped OWE resources could find using PSA-like contracts

advantageous. The governments could develop the new industry since costs are

falling, OWE is becoming cheaper than hydrocarbons, and subsidies are ending. This

research also demonstrates under what high-level conditions and contract terms OWFs

can be financially successful for both Contractors and Governments using PSA-like

contract terms, and what should be considered when negotiating.

CHAPTER 3

METHODOLOGY

Mixed Methods

This mixed method research follows a convergent design utilising two

interleaving exploratory phases. Since the 1990s, the importance of combining

qualitative and quantitative methods in social research has been recognised (Allen-

Meares & Lane, 1990). Mixed methods can aid in furthering research in business

studies because the type of data and analysis are different in the two phases and can

provide a broader or deeper understanding than quantitative or qualitative research

alone (Molina-Azorin, 2016; Kelle & Erzberger, 2004). In this research, neither the

quantitative nor qualitative approach alone would have been adequate to explore the

research questions fully.

In this concurrent triangulation approach (Cresswell & Cresswell, 2017), also

known as a convergent parallel approach, the two databases generated are not

comparable during analysis, but the findings are drawn together in the discussion

phase of the research. The strength of the approach is the resultant triangulation of the

conclusions, coupled with the time efficiency of parallel data collection. This approach

is clearly applicable to this research problem/opportunity. The quantitative and

qualitative phases were independent, answering two important but separate questions.

The quantitative phase examined whether the PSA-like fiscal terms work

financially/mathematically for OWE and provide positive returns to entice both

contracting parties to enter into the agreement. The qualitative phase investigated

whether all the terms, fiscal and non-fiscal, can be applied to OWE and what benefits

might accrue to the contracting parties besides a positive NPV. The knowledge

uncovered by the two phases was amalgamated into one conclusion on the applicability

of PSA-like terms for OWE. Arguments could have been made for a phased approach

in either combination. Finalising the quantitative research first could have provided

interesting information for the qualitative stage participants to discuss or conversely it

could have biased their opinions. Finalising the qualitative research first could have

provided the impetus for moving forward with the quantitative research – without

industry experts who see benefits to using PSA-like terms for OWE, the financial

outcome would have been a moot point. The drive for efficiency dictated that the

phases were pursued simultaneously; when there was a waiting period for one (e.g.,

waiting for Institutional Review Board (IRB) or scheduled interviews), the other

research went ahead, and when there was a deadline (such as an interview date), the

focus switched in order to conduct and transcribe the interview, and perform the first

level coding while the information was fresh in the researcher’s mind. This approach

had the advantage that the conclusions from one phase could not bias the other; both

were truly exploratory.

Researcher Credentials

The researcher has 30 years’ experience in the hydrocarbon industry and

although she has no direct career involvement in the renewable energy sector, she is

increasingly compelled to contribute to the urgent transition from the former to the

latter. She modelled PSA fiscal terms in Microsoft (MS) Excel and ran project

financial analyses for decision-making from 1998 to 2006 at four different O&G

companies: a super-major, two international independents, and a consultancy. The

researcher’s long experience with modelling contract terms and MS Excel will provide

a good foundation

for high-grade quantitative analysis. In the qualitative study, the researcher is a peer of

the participants and is skilled and knowledgeable about PSAs and their application in

O&G. From 2007 she was in management, corporate strategy, negotiations, and

mergers and acquisitions for two NOCs, which involved a detailed understanding of

contracts and their terms’ effects on the parties’ financial and non-financial positions.

During this management period, she gained practical skills relevant to research such as

interviewing and parsing themes from conversations to convert into corporate strategy.

The researcher’s long experience will assist in understanding the concepts

uncovered in the interviews. The researcher’s relatively recent interest in the wider

application of the terms to renewable energy will not be an impediment to

understanding as the participants are experts in O&G, not renewable energy.

Theoretical Foundations and Researcher Approach

The application of PSA-like terms to OWE has never been studied before (at least

nothing has been published). Thus, there is no established theoretical lens through

which to investigate. The research is exploratory in that the use of PSA-like terms for

OWE is a new concept. Thus, the epistemological approach to the research is

grounded theory. Whilst theory creation as such is not sought, if the quantitative

hypotheses are accepted and the qualitative research yields a practical application,

an existing concept (PSA terms) will be applied in a new context (OWE), which is

one use of grounded theory (Kelle, 2014). The research is not seeking to study or

explain an existing phenomenon but to introduce a new idea which can play a small

part in the transition from fossil

fuels. In the interviews, the initial discussion under RQ4 will serve as a springboard

for the discussion of RQ5, the new concept.

The researcher is a post-positivist in that she believes in and seeks explanations,

but also a pragmatist in believing there can be more than one interpretation and

explanation depending on the perspective of the viewer. Whilst post-positivism relies

on deduction, pragmatism embraces induction. Induction serves well in this

exploratory research. The researcher believes that human activity affects our world

for better or for worse; regardless of whether climate change was/is human-made,

constructive changes to our planet can be made in the future by humankind, by

changing our behaviour. The researcher believes an answer to the hypotheses can be

discovered empirically and although the participants may have differing opinions, the

quantitative simulation will reveal truths about the use of PSA-like terms for OWE.

Data Collection Schedule

This research study advanced through the process as required within the URI

dissertation framework. Namely, a dissertation committee was formulated by Spring

2023 – the committee heard/read the comprehensive examination. By late June of that

year, IRB approval was received for research on the American interviewees. By late

August the IRB approved research on the Norwegian interviewee.

The quantitative portion of this mixed method research commenced in June 2023

utilising data generated by simulation and therefore not requiring IRB approval. The

cash flow model and simulator had been built in 2022 and the selection of input

distributions was finalised in June 2023. The model could then be used at any time to

generate data. The data for the T-tests and multiple regression was collected in August

of the same year. The data for the scenario analysis was collected in September and

October and the qualitative interviews were conducted during July, August, and

September.

Data for the two types of research proceeded simultaneously. As mentioned

before, when there was a natural pause in one data collection effort (e.g., whilst

waiting for the second IRB approval), the other data collection could become the

focus.

Quantitative Study

Meta-modelling to Analyse Fiscal Regime Feasibility

Traditionally the O&G industry has used deterministic cash flow modelling

techniques to evaluate assets and projects, with decision trees and Swanson’s means

employed to assess risk (Researcher’s experience, 1992-2021). Papers are written,

entire conferences happen, e.g., Mian (2020), SPE HEES (1991 to 2016), to try and

shift the industry towards more sophisticated stochastic methods, yet companies’

formal decision-making processes remain simple, and/or not fit for purpose (Mian).

Even where stochastic analysis is seen, with input variables represented by

distributions, the use of formal statistical methods such as multiple regression is not

included (Researcher’s experience).

Kaiser & Pulsipher (hereinafter known as KP) introduced the concept of meta-

modelling to the O&G industry in their series of four papers from 2004 to 2005

(2004a; 2004b; 2005a; 2005b) and applied it to studies in the USA’s Gulf of Mexico

region (2004b) and Angola (2005b). The methodology has further been applied to

assets in China (Hao & Kaiser, 2010) and Nigeria (Adenikinju & Oderinde, 2009;

Ojaraida et al, 2018). As KP point out (2004a) there are many uncertainties involved

in evaluating an oil or gas field prior to an investment decision. These include

geologic risk, and production, price, cost, investment, technological, and strategic

uncertainties. A deterministic cash flow model, in which one fixed value is selected

for each input data point, cannot encompass these uncertainties. Analysis in the O&G

industry typically centres around a base-case outcome, compared to several “high”

and “low” scenarios in which one or more variables are set to higher or lower values

(KP 2004a). The meta-modelling approach goes beyond deterministic cash flow

analysis, using the model together with simulation to generate meta-data. This meta-

data undergoes multiple regression analysis to understand the interaction of the

variables (KP, 2005a). In their application of meta-modelling to the typical Angolan

PSA (Model I in their paper), KP (2005b) select seven parameters to be represented

by uniform distributions in the simulation: Production, Royalty Rate, CRC, Uplift

Rate (applied to capital costs), Tax Rate, Contractor Discount Rate and Government

Discount Rate. The output variables from the cash flow model are Contractor Take,

Net Present Value (NPV), and Internal Rate of Return (IRR). The regression

equations for each output variable give insight into the strength of the effect of a

change in each input parameter. In turn, this can inform the Government in setting an

appropriate fiscal policy for the hydrocarbon industry and can guide both Government

and Contractor in agreeing acceptable negotiable terms in the PSA.

This quantitative research used KP’s method with some changes to the input and

output variables. In KP’s three papers on PSAs (2005a, 2005b, and 2006), the myriad

fiscal terms to be found in the contract were simplified to four key predictor variables

which are considered in this research: Royalty, Cost Recovery Ceiling, Profit Oil

Split, and Tax. The fact that these four are the key components of a PSA is confirmed

by Tordo et al and Adenikinju & Oderinde. In addition, Uplift, a commonly used

fiscal term to encourage investors to new geographic, geologic, or technological areas

will be included, as KP did in their analysis of the Girassol field under an Angolan

PSA (2005b), for a total of five contract-term independent variables (IV). Profit Oil

Split is renamed Contractor Profit Split (CPS) for consistency with the OWE work to

follow. As offshore wind is a new technology and building OWFs under PSA terms

means bringing the farms to new geographies, an Uplift is considered an appropriate

inducement to investors. This research then advances the modelling technique by

applying it to an OWE project rather than the oil fields on which it has been used to

date.

Future studies could expand this predictor/IV set to include a further list of fiscal

terms that have a lesser effect on the project economic results.

Similar to KP’s method, Production (renamed Generation for the OWE model),

Price, and Contractor Discount Rate were non-contract term inputs/IVs represented by

distributions. Additionally, Government Discount Rate was included as the ninth and

final IV, as it is necessary for the calculation of Government NPV. Details of all the

distributions of these variables are given in the following sections.

This research goes beyond KP’s in a second way by looking at the contract terms

from the perspective of both contracting parties. Government NPV is included as an

output/dependent variable (DV), as it is the Government who would offer the lease

under these contract terms, so a positive financial outcome is critical for them.

Contractor NPV is the second DV, as a positive NPV is also critical for them, as the

investor. Future studies could expand this DV set to include other economic measures

such as Contractor IRR, Contractor Take (both of which KP studied) and Government

Take. The variable names, codes, acronyms, and types can be found in Table 1.

Definitions can be found in Appendix 2.

Variable Name Code Variable Type

Royalty Rate ry Independent

Cost Recovery Ceiling (CRC) crc Independent

Contractor Profit Split (CPS) cps Independent

Uplift Rate upr Independent

Tax Rate cit Independent

Average Price avpr Independent

Average Power Generation gen Independent

Contractor Discount Rate cdr Independent

Government Discount Rate gdr Independent

Contractor NPV crnpv Dependent

Government NPV gvnpv Dependent

Table 1: Variable codes, names, acronyms, and type

The cash flow model was built using MS Excel and screen shots of the tables and

worksheets can be found in Appendix 3.

The parameters for the distributions representing the stochastic variables are

detailed in the next sections of this paper. These are seen in action in Figure 27

(Appendix 3) on the right-hand side and feed the input variable cells in the same

figure

on the left-hand side. The output variables are also captured on the bottom right of the

same figure.

A snapshot of the cash flow model can be seen in Appendix 3, Figure 28, and

Figure 29 and Figure 30 show the cost input deck and the depreciation calculation.

A macro can be run to calculate the model as many times as desired, with

different values of the stochastic input variables selected by Excel for each trial. The

cash flow is recalculated, and the new output variables are displayed. For each trial,

the macro captures the input and output variable values as shown in the trial collection

data worksheet seen in Appendix 3, Figure 31. This is the raw data that can be

exported to run statistical tests or examined in scenario analysis. For each trial, this

worksheet collects the two dependent/output variables, Contractor and Government

NPV, the five contract-term IVs (Royalty, CRC, CPS, Uplift Rate, and Income Tax

Rate) and the four non-contract term IVs (Average Price and Average Generation

over the life of the project, and Contractor and Government Discount Rates).

Please note that the convention of M signifying thousands and MM signifying

millions has been used where applicable.

Calibration of Cash Flow

Model

The cash flow model was first calibrated by replicating KP’s analysis of a PSA

(KP, 2005a), applied to an oil field. The model file name is Economic Model - PSA for

OWE Kaiser Oil Inputs. The IV distributions in Table 2 were lifted from KP’s paper

(Model I) and input into this research’s model. The other model inputs (oil production

profile and cost profiles) can be found in Appendix 4.

Independent Variable Distribution Type Lower Bound Upper Bound

Price (Oil) Uniform $10 $30

Royalty Rate Uniform 10% 30%

CRC Uniform 10% 75%

CPS Uniform 10% 75%

Tax Rate Uniform 20% 40%

Contractor Discount Rate Uniform 15% 40%

Table 2: KP 2005a independent variables

The values for the distribution in KP’s research differ from the values used later

in this OWE research as the economic conditions were different in the O&G industry

two decades ago. However, for the purposes of calibration, they can be replicated. KP

had three DVs, NPV, IRR and Contractor Take, only one of which coincides with the

DVs in this research (Contractor NPV and Government NPV). This calibration

therefore uses Contractor NPV as the DV.

The model for this research will not match KP’s exactly. It is not known what

country’s PSA contract structure KP used as the framework for their model, nor what

accounting and tax laws they modelled, nor whether they included more minor PSA

terms such as bonuses, DMOs, training budgets, and others. For the sake of

parsimony, this research does not include these. However, the direction of each

multiple regression coefficient was expected to be the same and the magnitude of the

coefficients are expected to be similar relative to one another. The model was run

10,000 times and the data was collected to run a multiple regression. KP only ran their

model for 500 trials.

The regression model used matches KP’s for Contractor NPV, namely:

Contractor NPV = ß1 + ß2 * RoyaltyRate + ß3 *CostRecoveryCeiling +

ß4 * ContractorProfitShare + ß5 * TaxRate + ß6 * ContractorDiscountRate + ß7 *

AveragePrice.

Equation 1: KP and calibration exercise regression model

The expectation of the direction of the coefficients and KP’s coefficient values

are shown in Table 3. The constant, ß1 = -25.3 and R2 = 0.87

Coefficient Variable Expected Sign when

Contractor NPV is

Regressed on IV

KP’s Coefficient

Values

ß2Royalty Rate Negative -0.7

ß3CRC Positive 24.0

ß4CPS Positive 118.2

ß5Corporate Income Tax Rate Negative -80.8

ß6Contractor Discount Rate Negative -204.1

ß7Average Price Positive 3.7

Table 3: Expectation of regression coefficients and the values KP found

The following tests were satisfied:

Ratio of cases to independent variables: there are six IVs in each regression and 10,000

cases so that requirement is well satisfied.

Outliers: The DV was plotted against each of the six IVs (Appendix 5), and none

revealed outliers, so no data was removed.

Multicollinearity and singularity: the average Variance Inflation Factor was calculated

in Stata from the R2s. The result for Contractor NPV was 1.0000, which is less than

10, so multicollinearity is not an issue.

Normality: the plot of residuals for the DV, Contractor NPV, was generated in Excel

and the data were normally distributed. They can be seen in Appendix 6.

Heteroscedasticity & Linearity: Residuals plots for all IVs look linear. CPS shows a

hint of heteroscedasticity and Contractor Discount Rate shows a little more, but

neither are to a concerning degree, so the data is left as is. Plots can be seen in

Appendix 7.

No cases had missing data, n = 10,000. Table 4 displays the results for regressing

Contractor NPV on the six independent variables from the file Dataset1. The

regression was statistically significant, with R-squared = 0.83, F (6, 9993) = 8330, p <

0.001.

Table 4: Regression of Contractor NPV on six IVs

All six IVs had a significant effect on the DV with this research’s model, whereas

KP found that Royalty did not have a significant effect on Contractor NPV. They

found it did have a significant effect on the other DVs they measured, namely IRR and

Contractor Take, and the authors did not comment on this apparent discrepancy in

their discussion.

R2s were similar at 83% and 87% and the coefficients had the same direction and

relative magnitude in all cases except Royalty (Table 5). For example, both sets of

research found Contractor Discount Rate and CPS to have the most impact on

Contractor NPV. A larger sample size than 500 in KP’s research may have yielded

easier-to-explain results for Royalty Rate.

Coefficient Variable Values from this

Research

KP’s Coefficient

Values

ß2Royalty Rate -65.3 -0.7

ß3CRC 12.3 24.0

ß4CPS 84.4 118.2

ß5Corporate Income Tax Rate -52.4 -80.8

ß6Contractor Discount Rate -145.8 -204.1

ß7Average Price 2.6 3.7

Table 5: Comparison of regression coefficients from KP's and this research's models

Number of Trials

The calibration model sample size was 10,000 trials and the exercise was run ten

times to confirm that the sample size was sufficient. Table 30 in Appendix 8 shows the

constant and coefficients in each of the ten runs, from files Dataset1 through

Dataset10, and the means, standard deviations, and coefficients of variation.

The coefficient of variation (standard deviation/mean) is below 5% for all IVs in-

dicating stability of the outputs (DVs). The coefficient of variation for the constant is

higher at 14.4% indicating stability but of a lesser degree. A sample size of 10,000 tri-

als is sufficient for the research.

Parameters for Fiscal Contract Terms

The initial distributions for the variables Royalty, CRC, CPS, and Income Tax

Rate are based on current global PSA contracts from the confidential database

provided for this research by Wood Mackenzie (2022). Uplift tends to differ from

contract to contract and is not available by country in the database. Uplift will

therefore be modelled as KP did in their paper studying the Girassol field (2005b),

governed by a contract which includes uplift. The distribution is a uniform

distribution bounded by 30% and 50%.

The distribution for Royalty rates was identified from the mean of the mid-point

of each country’s oil and gas royalties, not including countries that do not have

Royalty in their fiscal regime. For example, [country name redacted1] oil Royalty

varies between 10% and 15% depending on the return of the project. The gas Royalty

also varies between the same points. The mean point for each fluid is therefore 12.5%

and the mean of the mid-points is 12.5%. [Country name redacted1] was removed as

an outlier having a Royalty mid-point of 50.00% compared to the next highest mid-

point of 20.00% (Figure 4).

1 Country name redacted due to NDA signed between the researcher and Wood Mackenzie

Figure 4: 118 Royalty rate mid-points (59 oil, 59 gas) illustrating two outliers

The distribution of the means of the mid-points over the remaining 58 countries

with PSAs with Royalties is shown in Figure 5. The mean is 8.50% and the mean

standard deviation of the two sets (oil and gas) is 3.97%. The range is 0% to 20%.

Figure 5: Histogram of Royalty rate mid-points (58 countries; 116 data points) illustrating normality

180160140120100806040200

60.00%

50.00%

40.00%

30.00%

20.00%

10.00%

0.00%

Mid-points of Royalty Rates, both fluids: Two Outliers

The distribution was subjected to the Kolmogorov-Smirnov test to check for

normality. The null hypothesis was that the data are normally distributed. N = 117 so

for an alpha of 0.05, the critical value is 0.1257. The calculated maximum was 0.0956

so the null hypothesis cannot be rejected, and the data are normally distributed. The

mean was not weighted by the number of contracts as that number is unknown – and

these input variables are merely starting points which will have a later adjustment to

optimise for an OWE regime. The value for the IV Royalty is bounded by 0% at the

low end of the scale and 100% at the high end, as neither negative Royalty, nor

Royalty above 100% have real-world meaning.

CRC, CPS, and Corporate Income Tax were calculated in the same way. No

outliers were found (Figure 6).

Figure 6: Scattergrams of the mid-point of each country’s CRCs, CPSs, and Income Tax for both fluid

types

The mean of the mid-points of the O&G CRCs is 66.94% and the standard

deviation is 15.70%. The range is 35% to 100%. The distribution is shown in Figure

Figure 7: Histogram of CRC rate mid-points (82 countries; 161 data points) illustrating a lack of

normality

The distribution was subjected to the Kolmogorov-Smirnov test to check for

normality. The null hypothesis was that the data are normally distributed. N = 162 so

for an alpha of 0.05, the critical value is 0.1069. The calculated maximum was 0.1121

so the null hypothesis is rejected, and the data are not normally distributed. This is in

part due to the large number of countries that have the CRC set at 100%. 100%

cannot be excluded in this analysis as it is a common value for the contract term. The

distribution for CRC for this research is therefore represented by a uniform

distribution bounded by 30% and 100%, the upper and lower bounds of the Wood

Mackenzie data.

The mean of the mid-points of the oil and gas CPSs is 50.85% (to the Contractor;

profit split to the Government is therefore 49.15%) and the standard deviation is

15.33%. The range is 17.5% to 95%. The distribution is shown in Figure 8.

Figure 8: Histogram of CPS Rate mid-points (82 countries; 161 data points) illustrating normality

The distribution was subjected to the Kolmogorov-Smirnov test to check for

normality. The null hypothesis was that the data are normally distributed. N = 161 so

for an alpha of 0.05, the critical value is 0.1257. The calculated maximum was 0.0749

so the null hypothesis cannot be rejected, and the data are normally distributed. The

value for the IV CPS is bounded by 0% at the low end of the scale and 100% at the

high end, as neither negative CPS nor CPS above 100% have real-world meaning.

The mean of the corporate income tax rates is 31.32% and the standard deviation

is 10.72%. The range is 12.5% to 65%. The distribution is shown in Figure 9.

Figure 9: Histogram of Income Tax Rate mid-points (80 countries; 80 data points) illustrating normality

The distribution was subjected to the Kolmogorov-Smirnov test to check for

normality. The null hypothesis was that the data are normally distributed. N = 80 so

for an alpha of 0.05, the critical value is 0.1521. The calculated maximum was 0.0996

so the null hypothesis cannot be rejected, and the data are normally distributed. The

value for the IV Income Tax Rate is bounded by 0% at the low end of the scale and

100% at the high end, as a negative Income Tax Rate or an Income Tax Rate above

100% has no real-world meaning.

Wind Generation Data

The latest advance in offshore wind turbine size came in early 2023 which saw

the construction of the first farm of 16 MW turbines in China (Lewis, 2023a). An

even larger 18 MW turbine was being prototyped in the same year and 2024 saw the

design of the first 20 MW turbine (Koragappa & Verdin, 2024). Slightly smaller 15

MW turbines first came online in 2022 and are to be installed offshore New York and

New

Jersey in the mid-2020s (Lewis, 2023b), so 15 MW turbines will be used for the

quantitative portion of this research.

Similarly, the average size of a large (over 500 MW) fixed bottom OWF has been

increasing year on year. Latterly in the last decade, from 2017 to 2019, such farms had

an average capacity of 703MW, as can be seen in Table 6. However, farms over 500

MW built in this decade are bigger, averaging 848 MW, also in Table 6.

Wind Farm Source Capacity

This decade: 2020 to 2022 Average: 848

Kreigers Flak Vattenfall (2021) 605

East Anglia 1 Buljan (2020) 714

Borssele 3&4 Power Technology (2021) 732

Borssele 1&2 Power Technology (2020) 752

Jiangsu Qidong Lewis (2021) 802

Triton Knoll Nye (2022) 857

Moray East Penman (2022) 950

Hornsea 2 Frangoul (2022) 1368

Late last decade: 2017 to 2019 Average: 703

Hornsea 1 Durakovic (2020) 1218

Walney Extension Unwin (2019) 659

Gemini Tsanova (2017) 600

Beatrice Power Technology (2022) 588

Gode Wind 1&2 Power Technology (2017) 582

Race Bank Ørsted (2018) 573

Table 6: Offshore wind farms that came online between 2017 and 2019 and their capacities

Planned wind farms that will follow from the 2022 California lease round, as

reported by Ayoub (2023) will be of an even larger 1 GW size, which will be common

by 2030. India (Times of India) and Brazil (Buljan, 2023) are leasing for future 1 GW

and 2-3 GW OWFs respectively. Should a developing country wish to design and

implement new contract terms, offer seabed for lease, and see wind farms built, it

would take time, so this research assumes that the farms under PSA come online in

eleven years, in 2035. Since the research anticipates operational farms in 2035, a 1

GW farm will be modelled. Given Australia’s ambition to have four regions, each

with 10 GW capacity, online by 2037, from a starting point of only one area leased in

2023, a 1 GW farm in 2035 in a country with no leases today does not seem overly

optimistic.

A 1 GW farm would require sixty-seven 15 MW turbines and would produce

5360 GWh/yr of power (calculated from Lewis, 2023b). This implies a capacity factor

(average power generated divided by peak capacity) of 61%. A 1 GW farm could

produce 1 x 24 x 365 GWh = 8760 GWh annually running at peak capacity; 61% of

that is 5360 GWh. NREL (2022a; can also be seen in Appendix 9) expects capacity

factors to average 50% (range of 46% to 54%) in 2035 for the best potential sites

(over 8m/s wind strength). Combining these data, since both sources are used in this

research, gives a capacity factor range of 46% to 61%.

At the early stages of evaluating offshore wind projects the industry normally

assumes a fixed generation profile for the life of the wind farm. It is not until the

construction or operation phase, when site-specific lidar data are collected, that the

profile would vary for the purposes of cash flow modelling (name withheld, 2023).

The profile for this research will therefore be homogenous year-on-year and will

follow a uniform distribution based on a 46% to 61% capacity factor.

Power Price Data

The wholesale power price is calculated from the Statista data set (Global Power

Prices, 2023), shown in Appendix 10. Only countries which have offshore wind

currently or have leased for offshore wind are included in the price distribution, as the

price for power in countries which do not have, or may never have OWE, are not

deemed relevant. These countries may have no shoreline (Rwanda) or may have such

large oil and gas reserves that power generation is cheap and heavily subsidised (UAE,

Russia, Saudi Arabia, Qatar, and Iran). In these latter countries, OWE would never be

competitive therefore contract terms are irrelevant, and they would not be an audience

for this research. Some countries may be future candidates for OWE (e.g., Italy,

Singapore, South Africa, Mexico, India, and Turkey), depending on many factors

including future power price. If the governments begin to offer leases the price from

Statista could be included in an update of this research.

The retail power price in the remaining countries is then converted to a wholesale

price based on the assumption that 56% of the retail price is the wholesale price (EIA,

2023). The mean of the resultant distribution is US$141.86/MWh and the standard

deviation is US$69.147/MWh. The range is US$44.8/MWh to US$268.8/MWh. The

distribution was subjected to the Kolmogorov-Smirnov test to check for normality.

The null hypothesis was that the data are normally distributed. N = 12 so for an alpha

of 0.05, the critical value is 0.1083. The calculated maximum was 0.9605 so the null

hypothesis is rejected, and the data are not normally distributed. The distribution for

power price for this research is therefore represented by a uniform distribution

bounded by US$44.8/MWh and US$268.8/MWh. Since the levelised cost of energy

from offshore wind is now dropping below US$50/MWh (IEA, 2020), this lower

bound seems realistic.

Wind Farm Cost Data

The cost inputs are taken from a dataset provided by NREL (2022a), hereinafter

referred to as ATB (annual technology baseline). Explanations and definitions of the

dataset can be found at NREL (2022b). The data used in this analysis can be found in

Appendix 9.

Construction Capital Expenditure

The costs used are 2032 forecasts for the construction of class 1 to 5 farms (where

class refers to the wind speed). Class 6 and 7 wind speeds were not included as they

represent lower wind speeds and projects less likely to be economic. It seems

reasonable to expect the first wind farms in new countries to be in the best areas. The

forecasts for 2032 are used as the contracts would be signed for construction

commencement in 2032 in order that the OWF be operational in 2035. As the contract

price would be fixed in 2032 it would be unrealistic to expect the 2033 and 2034

construction costs to fall in line with the ATB dataset forecasts.

Each class contains three cost estimates: advanced, moderate, and conservative.

This refers, among other things, to the turbine rating - the moderate, or middle, case

matches the 15 GW turbine assumption in this research, therefore the moderate case is

used in the analysis. The other cases could be used when analysing the robustness of

the project economics to changes in cost. The costs do not exhibit any clear

distribution, so a uniform distribution is used in this research where the upper bound

matches the class 1 cost estimate and the lower bound is the class 5 cost estimate.

The split of the total cost over the three-year construction time is 40:40:20 per the

ATB dataset and Appendix 9. The distribution is sampled once for a total capital cost

in 2032 and that total is split by the aforementioned ratio to divide it between the three

construction years 2032, 2033, and 2034.

Pre-construction Capital Expenditure

It is likely that funds are spent on management, studies, front-end engineering

design, and other tasks prior to construction. These amounts are not included in the

ATB dataset and are not reflected in this research. This approach is acceptable as they

would be small amounts compared to the funds required for construction, so they

would not influence the NPV greatly.

Grid Connection Costs

These are included in ATB estimate.

Financing

These are included in ATB estimate.

Operating and Maintenance Expense

The operating and maintenance (O&M) costs are also taken from the ATB

dataset. Again, the class 1 through 5 estimates are used, in the moderate case, to

create a uniform distribution from which to sample. ATB provides forecasts annually

through 2050, which are used in the analysis. The costs for the period 2051 to 2064

are not provided by ATB, so the 2050 costs are used. Arguments could be made for

increasing the cost by an inflationary measure - or for decreasing the cost due to the

learning curve, and those factors may cancel each other out, so no adjustments are

made. Since the O&M cost is only equivalent to 3% of the upfront capex amount and

is discounted by fifteen or more years from 2051 to 2064, it has little effect on the

outcome.

No decommissioning cost is included in the simulation as a large OWF has not

been decommissioned yet. Decommissioning of OWFs off the USA was modelled

over a decade ago by Kaiser & Snyder (2023) and the removal and site clearance

would be expected to cost US$115M to US$135M per MW. This implies an average

cost of US$125MM to decommission a 1GW OWF. However, given the fall in OWE

costs, estimates formulated a decade ago are potentially overstated. At that time,

OWFs were sized in MW, not GW, and economies of scale would apply. By 2019, the

largest OWF to be decommissioned was 10.5 MW, off Sweden (Adedipe & Shafiee,

2021). Finally, the cost will be discounted so heavily over thirty years that it will have

a minimal effect on the results, so it is omitted from the analysis.

The life of assets has been fixed for simplicity at 30 years (Sun et al, 2019).

Other Input Data

Discount Rates

Contractor and Government discount rates in the O&G industry are typically

different as the cost of capital for a government is usually lower than for an energy

company (Researcher’s experience, 1992-2021). The Contractor Discount Rate used

in this research is represented by a uniform distribution with an upper bound 10.88%

and a lower bound 7.04%, based on the weighted average cost of capital (WACC) for

O&G companies listed in Hegar’s analysis (2022). The distribution is based on the

100

mean of the O&G companies' WACC which was 11.94%. Inflation of 2.98% (based

on a US historical average from 1923 to 2022; Macrotrends, 2023) was deducted to

arrive at a real mean rate of 8.96%. The range for this analysis represents the mean

+/- two standard deviations (0.96; Hegar). The Contractor Discount Rate does not

vary with the risk of the project simulated in this analysis as the project is

hypothetical and does not exist in any specific location. Future analysis could be

refined to encompass varying Discount Rates based on project risk profile.

The Government Discount Rate uses the Contractor Discount Rate distribution

less 2%. The researcher knows from personal experience that NOCs typically use a

lower discount rate than publicly traded companies. From that experience the

researcher knows that some Governments may choose to use a 0% discount rate if

they are not contributing any capital to the project, then their NPV is an undiscounted

sum of cash flows. This research does not comment on the correctness or logic of that

policy. Since discount rates are typically held confidential, no literature has been

found as evidence of an appropriate rate for this analysis. Based on researcher

experience that governments using a non-zero rate tend to use a rate approximately 2%

lower than company rates, Contractor Discount Rate less 2% was selected. This results

in a uniform distribution bounded by 8.88% and 5.04%. Additionally, setting the two

parties' discount rates to different values will allow the researcher to gain insight into

how the parties might view a project differently and will demonstrate that the NPV of

the project can vary depending on how the cash flow is split between the parties.

PJM Peak NYMEX Futures Prices

2003-04-03 to 2019-11-29

180

160

140

120

100

0 500 1000 1500 2000 2500 3000 3500 4000 4500

PJM Peak NYMEX Futures

Prices

2003-04-03 to 2019-11-29

101

Inflation

The inflation rate used in the model is 0%, for the following two reasons. The

costs in the model are already stated in 2032 dollars and should not be subject to

inflation. Commodity prices such as oil, gas, and power are affected more strongly by

market forces than by inflation and it is hard to see an inflationary effect. For example,

the NYMEX power futures prices from 2003 to 2019 in Figure 10 show a downward

trend, though with an R2 of only 33% the trend is very weak, and it seems other factors

besides inflation are more strongly influencing the prices. 0% inflation has the added

benefit of assisting with ease of auditing calculations.

y = -0.0091x + 72.362

R² = 0.3299

Figure 10: PJM Peak NYMEX Futures Prices. Source: Capital IQ

102

Discounting Year

The research was carried out in 2023, so that year is used as the discounting year.

Depreciation

The PSA-user countries employ several different depreciation methodologies in

O&G (Wood Mackenzie); by far the most common is straight-line depreciation. Of

the 80 countries studied, 57 use straight-line depreciation. In this method, the same

amount is depreciated every year over the useful life of the asset or a defined number

of years. Of the other 23 countries, ten allow capital expenditures to be expensed and

seven use the unit of production method or declining balance method. Unit of

production is based on depreciating the costs in line with the proportion of total oil or

gas produced each year over the life of the field. Declining balance is an accelerated

depreciation method whereby more depreciation is taken in the early years of the

asset’s life. The remaining six countries use hybrid or other methods. The range of the

number of years over which the 57 countries allow straight-line depreciation is 2 to10

years, with the mean being 5.55 years and the most frequently occurring number of

years being five (30 instances), followed by four (12 instances) and ten (10 instances).

The exact tax rules would be set up as part of a government’s initiative to commence

leasing for OWE and would be studied in greater detail at that time to select the

optimal depreciation method. For this research, the most frequently occurring method

is modelled: 5-year straight-line depreciation.

Output Variables

Output variables are Contractor NPV and Government NPV as this study aims to

investigate both the Contractor and Government perspectives. These two output

103

variables are calculated by the cash flow model. Annual cash flow is calculated taking

into account gross revenue, based on power generation at that year’s prevailing price,

deducting costs, and then splitting that revenue between the Contractor and

Government based on the contract terms (Royalty, CRC, CPS, Uplift Rate and

Corporate Income Tax). Each of the two parties’ net cash flow is calculated and

discounted at the relevant discount rate. The discounted cash flow stream is summed

for each party’s NPV.

Economic Viability of OWE under PSA-like Terms

RQ1 investigated whether the wind farm was characterised by a positive NPV for

both the Contractor and the Government. The cash flow model was run 100,000 times,

the NPVs were captured for each trial in the file Results_20230901.csv, and the data

were loaded into Stata to run the T-tests to test hypotheses H1-1 and H1-2, namely:

H1-1: The mean Contractor NPV is positive.

H0 is that Contractor NPV is equal to 0.

Ha is that Contractor NPV is greater than 0, and:

H1-2: The mean Government NPV is positive.

H0 is that Government NPV is equal to 0.

Ha is that Government NPV is greater than 0.

Effect of Contract Terms and other Independent Variables on NPVs

The cash flow model was run 100,000 times, the NPVs were captured for each

trial in the file Results_20230901.csv, and the data were loaded into Stata to run

the multiple regression to test RQs 2 and 3, whose hypotheses are:

104

H2-1-1: Royalty Rate is expected to have a statistically significant negative

impact on Contractor NPV.

H2-2-1: Royalty Rate is expected to have a statistically significant positive impact

on Government NPV.

H2-1-2: Cost Recovery Ceiling is expected to have a statistically significant

positive impact on Contractor NPV.

H2-2-2: Cost Recovery Ceiling is expected to have a statistically significant

negative impact on Government NPV.

H2-1-3: Contractor Profit Split Rate* is expected to have a statistically significant

positive impact on Contractor NPV.

H2-2-3: Contractor Profit Split Rate* is expected to have a statistically significant

negative impact on Government NPV.

* The rate quoted is the proportion that goes to the Contractor. Government Profit

Split Rate is 100% minus CPS.

H2-1-4: Tax Rate is expected to have a statistically significant negative impact on

Contractor NPV.

H2-2-4: Tax Rate is expected to have a statistically significant positive impact

on Government NPV.

H2-1-5: Uplift Rate is expected to have a statistically significant positive impact

on Contractor NPV.

H2-2-5: Uplift Rate is expected to have a statistically significant negative impact

on Government NPV.

105

H3-1-1: Contractor Discount Rate is expected to have a statistically significant

negative impact on Contractor NPV.

H3-1-2: Contractor Discount Rate is not expected to have a statistically

significant negative impact on Government NPV.

H3-1-3: Government Discount Rate is expected to have a statistically significant

negative impact on Government NPV.

H3-1-4: Government Discount Rate is not expected to have a statistically

significant negative impact on Contractor NPV.

H3-2-1: The average Price is expected to have a statistically significant positive

impact on Contractor NPV.

H3-2-2: The average Price is expected to have a statistically significant positive

impact on Government NPV.

H3-3-1: Average Power Generation is expected to have a statistically significant

positive impact on Contractor NPV.

H3-3-2: Average Power Generation is expected to have a statistically significant

positive impact on Government NPV.

The regression models are as follows:

Contractor NPV = ß1c + ß2c * RoyaltyRate + ß3c * CostRecoveryCeiling + ß4c *

ContractorProfitShare + ß5c * TaxRate + ß6c * UpliftRate + ß7c *

ContractorDiscountRate + ß8c * GovernmentDiscountRate + ß9c * AveragePrice + ß10c

* AverageGeneration

Equation 2: Contractor NPV Regression Model

106

Government NPV = ß1g + ß2g * RoyaltyRate + ß3g * CostRecoveryCeiling + ß4g *

ContractorProfitShare + ß5g * TaxRate + ß6g * UpliftRate + ß7c *

ContractorDiscountRate + ß8g * GovernmentDiscountRate + ß9g *AveragePrice + ß10g

* AverageGeneration

Equation 3: Government NPV Regression Model

The following tests were satisfied:

Ratio of cases to independent variables: there are nine IVs in each regression and

100,000 cases so the requirement for twenty cases per IV is well satisfied.

Outliers: Both DVs were plotted against each of the nine IVs (Appendix 11), and none

revealed outliers, so no data was removed.

Normality: the plot of residuals for the DVs; Contractor and Government NPV were

generated in Excel and were normally distributed. They can be seen in Appendix 12.

Heteroscedasticity & Linearity: Residuals plots for all IVs as related to both DVs

appear linear. Plots can be seen in Appendix 13. No data exhibit the cone shape that

would indicate heteroscedasticity.

Multicollinearity: Variance Inflation Factor was calculated in Stata, and the mean

value for the nine IVs was 1.00 which is considerably less than 10 so multicollinearity

is not an issue. Results can be seen in Appendix 14.

No cases had missing data, n = 100,000.

107

Optimal Contract Terms

Multiple Regression allows us to look at the effects that each of the five contract

terms has on the NPVs over the whole distribution of each IV. In the dataset

Results_20230901, those ranges were rather wide as can be seen in Table 7.

Independent Variable Minimum Value Maximum Value

Royalty Rate 0.00% 27.30%

CRC 35.00% 100%

CPS 0.00% 100%

Tax Rate 0.00% 77.38%

Uplift Rate 30.00% 50.00%

Table 7: The minimum and maximum values for each independent variable

In a negotiation for a new PSA, these ranges would be narrowed quickly. The

Government would likely offer a set of contract terms, pinning the starting point for

each. As negotiations proceed, offers will be made around that starting point and not

utilising the whole distribution used in this research. Therefore, it is also important for

negotiators to have a methodology by which to evaluate the relative effects of the

fiscal levers within a narrower range. This methodology is scenario analysis, in which

up to eight IVs are held at a constant value and the remainder is/are allowed to take a

certain number of fixed values selected by the user. In these analyses the model was

run only 10,000 times for efficiency’s sake.

In this research, both stochastic and deterministic scenario analysis were employed.

108

Fiscal Terms in Depth

As a means to understand how the regression coefficient relates to the individual

effects of each fiscal term over its full range of values, the effect of each term on the

total NPV was studied. This allowed a deeper understanding of each contract term in

isolation – whether the regression coefficient represents the effect of the term at all

values it can take, or whether at the ends of the range the term behaves differently,

and the regression coefficient merely represents an average effect. The IV under

investigation was allowed to take on multiple different values across the same wide

range as seen in Table 7, while every other IV was held constant. A trial was run for

each IV value - and the values for IVs and DVs were collected. The total NPV was

calculated from Contractor NPV + Government NPV, and the change in total NPV

was plotted for each IV value. The dynamics within the cash flow model – the

interplay of the different fiscal levers – were scrutinised to uncover greater knowledge

about the dynamics of each fiscal term. For example, whether the total NPV rises

smoothly with an increase in Royalty rate or whether it rises in “stair steps” caused by

an event occurring in the cash flow model.

Optimising Fiscal Terms

This second type of scenario analysis aimed to discover if a symmetrical change

in a fiscal term IV yields a symmetrical change in the DV, as regression would imply.

For example, a regression coefficient of -10 for the IV Royalty Rate when regressed

on Contractor NPV implies that for a 1% rise in Royalty Rate, the NPV decreases by

10 and for a 1% fall in Royalty Rate, the NPV increases by 10. Four values were

109

selected for each IV to yield three resultant differences between NPVs. Table 8

contains the levels of each IV that were tested.

Independent Variable Mean of distribution

(for reference only)

Levels tested in analysis

Royalty Rate 8.5% 8%, 10%, 12%, 14%

CRC 67.5% 60%, 65%, 70%, 75%

CPS 50.85% 48%, 50%, 52%, 54%

Uplift Rate 40% 38%, 40%, 42%, 44%

Income Tax Rate 31.32% 28%, 30%, 32%, 34%

Table 8: The levels of each fiscal term / IV tested in the optimisation analysis

The total NPV results were analysed for symmetry with regards to rises and falls

in IV values and the Contractor NPV changes were compared to the Government NPV

changes. A comparison was also made, for the sake of interest, between the results of

this analysis and the regression coefficients.

Following this, an example is given, simulating a real negotiation, of how a

change in one fiscal term can be compensated by changes in others from one party’s

perspective, for example, if the Government suggests an increase in Royalty,

damaging the Contractor NPV, how the Contractor might request changes to other

terms to compensate. Both parties’ NPVs were considered – if both parties are better

off after a change in terms, the new set of terms may be the optimal set.

Stochastic Analysis for Negotiation

Having gained a deeper understanding of the behaviour of the fiscal terms and

having found a preferred set of terms during negotiation, the final step, which

combines the two latter techniques, is to re-introduce distributions for the IVs whose

110

values are

111

unknown at the time of negotiation. These are Price, Power Generation, and Cost.

Through an example, this research illustrates the kind of output a decision-maker

could employ prior to committing to the contract. The fiscal term IVs are set to their

average levels as if those were the currently agreed values in an active negotiation.

The distributions used in the multiple regression are re-introduced for Power

Generation, Price, Cost, and Discount Rates. 10,000 trials were run and the

distribution of NPVs for each party is illustrated.

NPV Sensitivity to Project Characteristics

Whilst Price, Generation, and Cost are not negotiated variables, a change in any

of them could render an otherwise economic project unviable, so they are also

important variables for both contract parties to examine. Sensitivity to project

characteristics such as these gives information about the risk profile of the project. It

also guides the Contractor as to which decisions have more impact on the outcomes

than others. For example, if the OWF is highly sensitive to Cost, the most effort

should go into lowering both cost and its variability. If the project is insensitive to

Generation, less time might be put into the siting of the turbines.

In order to study the effect of these variables on the applicability of the contract

terms, scenario analysis was again used. The five fiscal levers were held steady at their

mean values and the price, generation, and cost were set to three levels. Those levels

were: mean + 20%, mean, and mean -20%. 10,000 trials were run each time and the

mean values for IVs and DVs were collected.

112

Qualitative Study

Research Design

The purpose of the interviews was to elicit opinions and ideas as to the

applicability of PSA-like contract terms to OWE from current or former users of

PSAs.

This understanding complements the findings from the quantitative phase; under

which circumstances PSA-like terms can be applied to OWE. Additionally, that phase

studies only the financial contract terms and does not quantify the non-financial terms

discussed in Chapter Two. Discussions with industry professionals who are experts in

PSAs can test the applicability of these other benefits to OWE.

Research Location

The participants were current and former employees of O&G companies and were

located in Texas, West Virginia, Connecticut, and Norway. Due to cost and time

constraints, the setting was seven synchronous virtual video meetings via MS Teams

or Zoom. There was very little control over the physical location of the participants,

being remote, but they were all in their offices or residences. The location of the

participants does not affect their knowledge of the subject matter; all have worked in

international settings whilst exposed to the PSAs they have experienced.

Population

The study population was current and former O&G industry employees with

significant experience in PSAs and at least basic knowledge of OWE. This population

is able to talk in detail about the facets of PSAs and the advantages and disadvantages

to both contracting parties. With their long experience in the industry these experts

113

typically have opinions on the transition and renewable energy, and some experience

with offshore wind.

Sample

The sample of seven interview participants was selected by non-probability

methods i.e., purposive sampling. It was not possible to ensure the representation of

the population of all O&G employees with PSA experience due to the purposive

sampling through personal relationships. However, the participants were carefully

selected to represent different company types within the industry – IOCs (two), NOCs

(three), and industry consultancies (three). They were selected for their experience

working in detail with PSAs. The study has limited generalisability based on the small

sample size.

The participants were recruited through personal relationships with the

researcher. One was an industry peer, three were former consultants, and four were

ex-colleagues. No research was done until the research proposal was approved, and

the IRB requirement was satisfied, but initial contact to canvas willingness was made

prior to IRB approval in June 2023. IRB approval was attained in two tranches: first

to interview American citizens, and second to interview Norwegian citizens

(ultimately only one agreed to participate).

The sample demographics do not mirror society as a whole but are fairly

representative of white-collar O&G workers and retirees, who would have experience

with PSAs. The sample is biased towards Americans due to the simpler IRB approval

process in which cross-cultural considerations are not relevant. Sample demographics

can be seen below in Table 9.

114

Participant Age Race Gender Education

Level

Income Nationality Company

Type

1 Over

White Male Masters 100-

300k

American IOC

2 40-

White Male Masters 100-

300k

American Consultancy

3 40-

White Female Masters 100-

300k

American Consultancy

4 Over

White Male Bachelors Retired American Consultancy

5 Over

White Female Bachelors Retired American IOC

6 40-

Asian Male Doctorate 100-

300k

American

Malaysian

NOC

7 40-

White Female Masters 100-

300k

Norwegian NOC

Table 9: Demographics of interview participants

Data Collection

Verbal statements (and their accompanying non-verbal cues) were collected

through semi-standardised interviews, recorded with MS Teams or Zoom. The RQs

could not be answered through observation. RQ6 could have been investigated

through documentation, as it is a well-studied area, but the research would have

115

lacked opinions and discussion of a new concept (RQ7) had it been performed

through studying historical documents.

Pilot Interviews

Prior to IRB approval, two pilot interviews were conducted to test the questions.

At that stage, the participants were to be Government or Energy Ministry employees.

The pilot interviewees were industry ex-employees posing as Government employees

to test the questions. The questions underwent revision after each pilot until finalised.

Although the population and sample changed between the pilot interviews and IRB

approval, the questions were still relevant, with only minor wording changes for the

new audience.

Standardised/Structured Interview Protocol

An interview protocol is more than a simple list of questions; it outlines the

procedure of the interview and includes a script for the interviewer. It prompts the

interviewer to provide an introduction, to collect informed consent, and serves to

remind the interviewer the key demographic and research information to collect

(Jacob & Furgerson, 2012). The interview script can be found in Appendix 15. The

interviews were focussed in order to elicit deep reactions and context, and questions

were open-ended to spark discussion and idea generation. An introduction was given,

followed by the protocols (informed consent, confidentiality, data privacy and

sharing, the research process, and an offer of the research conclusions) such that

consent was achieved. Demographic questions followed, then the discussion

questions were asked. A survey would not have been appropriate for this exploratory

research (Jain, 2021), despite the fact a survey would have been quicker than

116

interviewing and transcribing multiple interviews. Nor would interviews with closed-

ended questions and a fixed script have been appropriate as they generate numeric

data rather than eliciting the participants’ experiences with PSAs (Ivankova &

Cresswell, 2009) which is required for Grounded Theory. The questions were focussed

but gave the participants room to be creative, to consider new ideas as they occurred,

and to jump between topics to follow their mental flow.

Response Recording

The researcher did not take notes during the interview in order to focus on the

conversation and listen intently. Since the interviews were recorded (and a backup

audio recording was made simultaneously with a hand-held dictaphone), notes were

not required during the interview. Additional follow-up discussion could have taken

place if information needed to be expanded and/or clarified as appropriate and

relevant. Initial impressions were noted immediately after the meeting.

The researcher was the interviewer, and for three of the seven interviews, a

colleague was present to confirm the researcher’s understanding via a second set of

eyes and ears. The debrief after those interviews confirmed the researcher had

correctly interpreted the data and highlighted any instances in which the researcher

could have led the direction of the conversation. The interviewer was mindful to listen

rather than lead the conversation, and speak only to provide support, move the

conversation, or inquire in more depth when merited. There were no such instances

of leading in the three observed interviews. There were no instances in which the

interview could not be conducted in English.

117

There was no time limit imposed on the interviews and six of the seven were

completed within 60 minutes. Transcriptions are available in raw, unedited form, and

in a sanitised form to assist clear understanding of the participant’s meaning.

Data Transcription

Standard orthography was employed during transcription – correct English

spelling and punctuation - with the addition of colloquialisms and industry-specific

terms where merited. Non-verbal/linguistic cues – paralinguistics, prosody, and

extra-linguistics were captured for all interviews but only transcribed for one

interview – the additional captured data did not add any useful information to the

coding process, so non-verbal/linguistic cues were not transcribed for the remaining

interviews. Transcription was in Otter.ai and each interview received three iterations

of listening to the raw audio whilst reviewing the automatic transcription to ensure

the most faithful transcription possible. In one case, there were unintelligible words

captured and noted in the transcription as “[unclear]” but in the researcher’s opinion

this did not dilute the participant’s meaning. In three cases, a second, sanitised

transcription was made. Stuttering and repeated words threatened to make coding

difficult, so those repetitions were deleted. The deletions did not diminish the

participants’ meanings.

Data Analysis

The approach to the qualitative data analysis is guided by the Straussian school

of Grounded Theory. Grounded theory was derived from pragmatism (Strauss &

Corbin, 1990) but the Thematic Analysis method that was employed is also

considered applicable to post-positivism (Clarke & Braun, 2017), the two stances of

118

this researcher. As Strauss permits, the literature review was performed and provides

knowledge about many topics adjacent to the research question, such as renewable

energy, offshore wind, hydrocarbon contract terms, and current trends in leasing,

subsidisation, and costs in OWE. It does not provide any background on the research

question itself, which has never been asked before.

Thematic Analysis is a flexible approach that can be applied to many data

sources, including interviews, and any dataset size, from a single case study to large

interview fields (Clarke & Braun). It is well applicable to this dataset of seven

interviews where themes are sought.

The Straussian approach calls for three stages of coding: open, axial, and

selective (Strauss & Corbin).

Open Coding

The first interview was coded immediately after transcription, employing open

coding. The interview questions were adjusted in very minor ways (wording only) to

function better for subsequent interviews. The second to seventh interviews were

coded shortly following transcription and the codebook was expanded as new themes

emerged. By the sixth interview saturation was being reached in the sense that new

topics were not being introduced, and the seventh confirmed concepts but did not add

new ones.

The concepts were organised and coded using NVivo software. Each transcript

was coded twice at the open coding level – once for O&G questions and once for

offshore wind questions. Each transcript was imported twice, labelled O&G or OWE,

and coding for O&G was performed on the O&G cases, whereas coding for the

119

transition and offshore wind was executed on the OWE cases. This was so that the

two intents can be kept separate when running NVivo queries – for example, Cost

Recovery was a common theme in both discussions and the researcher wished to be

able to query it as a part of O&G alone, and as a part of OWE alone, to see if the

changing context made a difference to how the participants spoke of the theme.

No prior coding of interviews on the topic of PSAs has been identified in the

literature so the codebook is original.

Memoing was used to capture the handling method of each transcript and any

comments of the observer. The raw audio and video were referred to in a few

instances to confirm or triangulate what was being said with non-verbal reactions and

ensure correct interpretation. Consistency was a concern prior to the research as it

used open-ended discussions – the concern was if the first interviews revealed ideas

worthy of exploration subsequently, but not envisaged at the pilot study stage or in

the question script. This concern did not come to fruition. No concepts emerged from

the interviews that caused the need for subsequent script changes. The only script

changes were to improve the clarity of the questions.

The transcripts were studied for the advantages and disadvantages of using PSAs

for O&G leasing to contracting parties (RQ6). Some advantages (or disadvantages)

affect both parties, and in other cases an advantage to one is a disadvantage to the

other. The transcripts were further studied for the participants’ opinions on the

transition and OWE, and how the features of the PSA could be applied to offshore

wind (RQ7). Each participant discussed the advantages and disadvantages they could

see from using a similar leasing system for offshore wind. Given the inductive coding

120

approach, data could be coded to several codes simultaneously depending on whether

it touched multiple themes.

After the seven interviews were completed, constant comparison was employed

(Strauss & Corbin). Each coded transcript was reviewed to ascertain if early

interviews should be coded to new codes that arrived later. Approximately fifteen

data points were re-coded or coded to additional codes.

The interpretation is empathic – the researcher did not seek to impose ideas on

the participants and given that she may be seen as an authority figure, care was

exercised. Suspicious interpretation was available to be employed to combine

responses containing contradictory ideas or responses based on incomplete or

incorrect knowledge. This only happened in one instance, where a respondent was

working on an incorrect assumption about OWE.

The code book as it stood after this phase can be seen in Appendix 17.

Axial Coding

In the second level coding, inductive methods were used to seek patterns in the

data; similarities, differences, frequency, sequence, correspondence, and causation

(Saldana, 2021). Despite the importance Strauss & Corbin place on similarities and

differences, they were not under investigation per se, as the goal is not to compare

opinions but to build a theoretical concept from multiple participants. Differences

provide the breadth of concepts and similarities provide the importance of a concept

among the whole. The frequency with which each concept arises was of interest, as

it could indicate importance. An idea may be mentioned multiple times by a few

participants or mentioned a few times by all participants. The sequence in which

121

participants raised topics may be of importance – a topic that is raised first by several

participants may be a key concept. The segue between topics may be interesting – or

may just indicate the participants’ thought patterns, for example, perhaps they are

thinking through the contract in page order as they speak. Sequence may also indicate

linkage between topics, for example, Cost Recovery and Profit Split are often spoken

of in the same breath as they are key differentiating characteristics of a PSA.

Causality was not a key concern in this research as an explanation of a behaviour was

not sought. The goal was the generation of key themes about a business proposition.

Explanatory power is important – why a certain concept is important to the

participants, and to the research question, but this research does not seek to identify

causes for the existence of that concept. The fact it is a characteristic of a PSA is

accepted by this research.

Although induction is complex and risks lacking focus, with the small number

of interviews to handle, it was manageable.

In the axial coding stage, codes were merged and moved as needed. Codes

related to the order of participants’ responses were added. In the open coding, the

researcher sought concepts and had not yet focussed on the importance of the

concepts. The code book as it stood after this phase can be seen in Appendix 18. A

coding query was next run to ensure all data had been coded to at least one code. The

few statements that had been missed that were not simply demographic responses or

punctuation were coded.

122

Selective Coding

In this final stage, the codes were further refined and pulled together to create

the over-arching story (Strauss & Corbin). The codes are no longer “buckets” of data

but systematically relate to one another. Where there were gaps in the story, the

transcripts were searched for answers.

The final code book be seen in Appendix 19.

Querying the Data

NVivo standard queries such as Word Frequency, Word Trees, Coding Query,

and Matrix Coding Query were employed to investigate the data, as well as

mapping the codes in PowerPoint, and analysing the data simply by reading and re-

reading. The use of these tools can be seen in the Results and Discussion section.

Methodological Integrity

In the quantitative study, the internal validity of the cash flow model used for

simulation was achieved through an audit of the model from both the major professor

and a second committee member, and through the replication of KP’s 2004 study.

The regression coefficients achieved by KP and in this research are all in the same

(expected) direction, and of similar relative magnitudes (please refer to Table 5). The

coefficients will never exactly match because KP provided neither their model nor

their detailed accounting assumptions. The model used in this research will differ,

though not significantly. Since neither KP’s nor this model are intended to replicate

a specific country’s PSA, they pull on the most commonly found contract terms, but

the researcher and KP may not have selected the same exact terms.

123

In addition, to ensure internal validity, the inputs into the model (please refer to

the section on meta-modelling) were selected from credible sources such as Wood

Mackenzie, the IEA, NREL, and Statista to ensure that conclusions drawn from the

model’s output can be relied upon.

Quantitative results are not generalisable outside the range of the input values.

However, external validity is achieved for a specific situation that matches the input

ranges used. A country which both has the wind potential to match the simulated

OWF at the cost ascribed to it and can sell the power within the price range simulated,

could look to this research’s conclusions to guide on applicability of PSA-like terms.

However, for a power price outside the range used, or a lower capacity factor than

modelled affecting the wind farm operation, a different data set would be collected

from the simulation, and results and conclusions may change. Before investing in

projects such as these, the best available inputs should be collected for the specific

context. For example, should a government in Nigeria be interested in these contract

terms, they must adapt the wind farm costs, power price, and generation profile to

local conditions.

Considering reliability, it is not possible to replicate these results exactly. The

simulation involves calculating the cash flow anew with each trial and the chance of

repeating the exact set of 100,000 or 10,000 data points is extremely small. For

example, three consecutive runs of 10,000 trials each yielded the following mean

NPVs in Table 10; the total NPVs are similar though the split between parties varies:

124

Mean of 10,000

NPVs

Contractor NPV Government NPV

Total NPV

Set 1 US$340MM US$1213MM US$1553

Set 2 US$274MM US$1313MM US$1587

Set 3 US$266MM US$1277MM US$1543

Table 10: Mean NPVs over three sets of 10,000 trials

Additionally, confidence can be placed in the T-tests with P-values of 0.0000

and in the multiple regressions with R2s of 84.41% and 90.16%. A sample size of

100,000 trials further improves the chance of avoiding bias.

Assessing validity in qualitative research is a debated topic. Current thinking is

that a single prescribed method to assess validity may in fact prevent quality analysis

due to the diverse nature of qualitative research (Welch & Piekkari, 2017). Further,

Hayashi et al (2019) recommend a processual rather than a traditional quantitative

approach to assessing validity. This means that the quality of the research is visited

and revisited throughout the work rather than by an ex-post test. For this reason, the

calculation of Cohen’s Kappa was not attempted, as the researcher feels it would be

fraught with errors given the lack of availability of a co-researcher with in-depth PSA

knowledge or a participant with a working knowledge of coding and NVivo.

Welch & Piekkari suggest the use of multiple sources of data and a chain of

evidence to demonstrate construct validity. Concerning the former, seven participants

were interviewed, and common themes quickly emerged. Saturation was achieved by

the sixth interview, indicating between-participant agreement. Data credibility was

achieved within-participants through triangulation with non-verbal cues where there

125

was any doubt in the researcher’s mind as to the intent of a participant’s comment.

The interview observer was able to confirm the researcher’s understanding of the

participant and that she did not lead the participants or in any way bias the responses.

During interviews, the researcher was careful to listen and only direct the

conversation to the extent that extra detail or explanation was required, or to

encourage a participant to examine all the issues and not get overly focussed on one

contract term. The researcher was aware that she could be biased toward the research

topic (PSA-like terms being applicable) and did not want participants to be swayed,

so she limited her verbalisation to that strictly necessary to keep the interview moving

and on topic. As a post-positivist, validity through the lenses of the participants and

people external to the study can be achieved through member and peer checks

respectively (Creswell & Miller, 2000). As such, a peer check was performed, which

confirmed that the findings and conclusions could feasibly have been drawn from

interviewing O&G professionals. The demographic data of the peer check participant

and some of the comments made in the interview can be found in Appendix 16. The

process used to interview the peer check participant was identical to that employed

in the seven interviews.

Regarding the process used, the researcher maintained multiple, carefully

organised copies of data to ensure no loss. Transcriptions are available in raw,

unedited form, and in a sanitised form to assist clear understanding of the participants’

meanings. Codebooks at the end of each of the three stages of coding are captured and

available in Appendices 17 to 19.

126

Internal validity was achieved through the three-stage coding process, extracting

themes and revisiting transcripts coded early as the codebook advanced. The

questions were focussed but gave the participants room to explore any topic or

tangent. There was no time limit set on the interviews and six of the seven were

completed within 60 minutes. A member check interview was not possible due to the

lack of availability of participants.

Transferability, or Welch & Piekkari’s last category, external validity, is implied

by the quick pathway to saturation in coding. Within six interviews the major themes

had been elicited and repeated by several or all participants. Since these participants

have global experience, covering many of the 82 PSA-using countries (Wood

Mackenzie), their comments can reasonably be assumed to be applicable to any PSA-

using country, and depending on the context, could well apply to any country, and

even to other renewable energy sources. Replicability of results is also more likely

following this quick pathway to saturation. Since this is the first work on the

applicability of PSA- like terms to OWE, there is no extant literature to which to

compare the results.

Confirmability was achieved by the fact that similar opinions came from

participants from two countries (USA and Norway) and from three company types

(NOC, IOC, and consultancy).

The qualitative research could be strengthened by interviewing governments or

energy ministries, to get the perspective of both parties to the contract. Interviewing

NOC employees goes a small way to achieve that, hence the selection of two

participants who work for NOCs. A bigger sample size would also add richness, but

127

saturation was reached with the seven interviews performed. Participants from more

than two countries may also yield more richness.

Resources Utilised

MS Excel was used for the simulation and Stata was used for the statistical

analysis. The latter’s licence required funds allocated to the URI DBA programme.

NVivo was used for the qualitative analysis, again, the licence required funding.

Otter.ai was used for transcription, the licence was kept on a month-to-month basis to

preserve programme funds. The literature review was conducted using Google

Scholar, the URI library, DBA class material, other web-based resources, and in some

cases direct contact with authors. The researcher’s database of 362 articles was held in

MS Excel and MS Word. The dissertation was written in MS Word and presented

using MS PowerPoint.

128

CHAPTER 4

FINDINGS

Economic Viability of OWE under PSA-like Terms

The results of the T-tests for the null hypotheses that Contractor NPV and

Government NPV (variables crnpv and gvnpv) are equal to zero are in Table 11. The

sample size is 100,000 trials and each test has one variable; degrees of freedom are

99,999.

Table 11: One-sample T-test results for the variables crnpv = 0 and gvnpv = 0

129

The first test, for Contractor NPV, indicates that the underlying mean is not equal

to 0 with a significance level of 0.0000. The test indicates that the NPV is greater than

zero with a significance level of 0.0000. The mean NPV for the Contractor is

US$234MM implying that on average, the rational Contractor would participate in this

project. The 95% confidence interval (CI) for the mean is US$232MM to US$236MM

– a narrow range – not surprising given 100,000 data points. The CI for a 500-point

data set such as KP analysed would be much wider. This CI means there is a 95%

chance the true mean falls in this range. However, the standard deviation of

US$302MM implies that there are instances in which the project does not yield a

positive NPV, in which case the rational investor would not move forward with the

project. The mean is rather variable, with the magnitude of one standard deviation

being higher than the magnitude of the mean itself, despite the confidence in the

mean’s accuracy. 31.8% of cases will result in an NPV more than one standard

deviation from the mean, which equates to a negative NPV in the case of mean minus

one standard deviation, 15.9% of the time. In fact, since the mean is US$234MM, the

NPV will be negative at minus 0.77 standard deviations, which will occur more than

15.9% of the time. For this sample, in 22,137 of 100,000 trials (22%), the Contractor

NPV was negative. However, with a positive mean NPV, the only information known

at the time of deciding on the project, the rational decision would be to enter into the

contract, in the absence of a better project competing for financial or human capital.

The second test, for Government NPV, indicates that the underlying mean is not

equal to 0 with a significance level of 0.0000. The test indicates that the NPV is

greater than zero with a significance level of 0.0000. The mean NPV for the

130

Government is US$2,046MM with a standard deviation of US$695MM implying that

it is unlikely that the project does not yield a positive NPV. In this sample, the

Government NPV was not negative in any of the 100,000 trials. A rational

Government investor would participate in this project. The CI for the mean is again a

small range, from US$2,042MM to US$2,051MM.

In fact, when the total NPV is taken into consideration, no projects in this sample

yielded a negative total NPV. The lowest total project NPV was US$343MM, split

such that Government NPV was US$697MM and Contractor NPV was US$-354MM

(a Contractor would not approve this project if it had this data ahead of time). The

highest total project NPV was US$5,816MM, split such that Government NPV was

US$5,473MM and Contractor NPV was US$343MM.

Pursuant to the T-tests, the null hypotheses associated with H1-1 and H1-2 were

rejected and both Government and Contractor means are statistically greater than zero

(RQ1). It can be concluded that under these contract terms, discount rates, generation,

price, and cost conditions, PSA-like terms can be used for OWE. Both contracting

parties will statistically realise a positive NPV and should rationally be willing to

move forward with the project, in the absence of another project competing for

financial or human capital.

Effect of Contract Terms and other Independent Variables on NPVs

Multiple regression was used to predict whether the IVs Royalty, CRC, CPS,

Uplift Rate, Tax, Contractor Discount Rate, Government Discount Rate, Average

Power Generation, and Average Power Price significantly predicted Contractor NPV

131

and Government NPV. The results can be found in Table 12. The sample size was

100,000 trials and the degrees of freedom were 99,990 given nine IVs and one DV per

test. The dataset used was Results_20230901.

Variables Contractor NPV (crnpv) Government NPV (gvnpv)

Royalty Rate (ry) -1,252.4383*** 1,760.8105***

Standard Errors [9.639] [17.666]

Cost Recovery Ceiling (crc) 289.4029*** -323.9299***

Standard Errors [2.007] [3.678]

Contractor Profit Split (cps) 1,024.4464*** -1,596.9042***

Standard Errors [2.506] [4.593]

Tax Rate (cit) -1,022.8288*** 1,580.2223***

Standard Errors [3.528] [6.466]

Uplift Rate (upr) 208.5613*** -305.2975***

Standard Errors [6.520] [11.950]

Contractor Discount Rate (cdr) -13,552.3872*** 68.19

Standard Errors [33.946] [62.216]

Government Discount Rate (gdr 11.8199 -45,264.7767***

Standard Errors [33.894] [62.119]

Average Generation (gen) 0.2417*** 0.6136***

Standard Errors [0.001] [0.002]

Average Price (avpr) 7.2612*** 18.3341***

Standard Errors [0.038] [0.070]

Constant -1,179.3988*** -78.2909***

Standard Errors [9.190] [16.844]

Observations 100,000 100,000

R-squared 0.8441 0.9016

*** p<0.01, ** p<0.05, * p<0.1

Table 12: Multiple regression of Contractor NPV and Government NPV on nine IVs

The overall regressions were statistically significant (R2 = 0.8441 for Contractor

NPV, F(9, 99990) = 60,149.79, p = 0.0000) and R2 = 0.9016 for Government NPV,

F(9, 99990) > 99999.00, p = 0.0000).

It was found that Royalty, CRC, CPS, Uplift Rate, Tax Rate, Contractor Discount

Rate, Average Power Generation, and Average Power Price significantly predicted

Contractor NPV. See Table 12 for β-values (coefficients), p = 0.0000 in all cases.

132

It was found that Royalty, CRC, CPS, Uplift Rate, Tax Rate, Government

Discount Rate, Average Power Generation, and Average Power Price significantly

predicted Government NPV. See Table 12 for β-values, p = 0.0000 in all cases.

The null hypotheses associated with hypotheses 2-1-1 through 2-2-5 and 3-1-1

through 3-2-2 were rejected following the multiple regression analysis. The IVs listed

above associated with those hypotheses have a significant effect on NPV. Note that

the null hypotheses associated with H3-1-2 and H3-1-4 were also rejected; the

expectation was that no significant effect would be found. Indeed, it was found that

Government Discount Rate did not significantly predict Contractor NPV (β =

11.8199, p = 0.727) and that Contractor Discount Rate did not significantly predict

Government NPV (β = 68.19003, p = 0.273). This is not a surprising result as

Government Discount Rate is not used in calculating Contractor NPV, nor is

Contractor Discount Rate used in calculating Government NPV.

The regression models are tabulated below in Table 13. Note that if using the

models for prediction purposes all IVs commonly expressed as percentages (all except

Average Price and Average Generation) must be expressed in decimal form.

133

Coefficient / Constant Contractor NPV Government NPV

Constant ß1c: -1179.99 ß1g: -78.29088

Royalty ß2c: -1252.438 ß2g: 1760.81

CRC ß3c: 289.4029 ß3g: -323.9299

CPS ß4c: 1024.446 ß4g: -1596.904

Tax Rate ß5c: -1022.829 ß5g: 1580.222

Uplift Rate ß6c: 208.5613 ß6g: -305.2975

Contractor Discount Rate ß7c: -13552.39 N/A

Government Discount Rate N/A ß8g: -45264.78

Average Generation ß9c: 0.2417061 ß9g: 0.6136451

Average Price ß10c: 7.261181 ß10g: 18.33407

Table 13: Coefficients to be used in forecasting and analysis

Sense-check of Regression Model

In order to test the regression model, consider a case with the inputs in Table 14

below (column “Value for Test”). These IV values are close to (rounded up or down)

the mean values for each variable in the dataset Results_20230901. These values have

no specific meaning, they are just selected for the convenience of double-checking the

regression model before proceeding with the analysis. If the DVs are calculated using

the regression model, Contractor NPV would be US$151MM and the Government

NPV would be US$1998MM. These are well aligned with the mean NPVs of

US$151MM and US$1980MM respectively from the same dataset

(Results_20230901), and the researcher is comfortable that the regression equations

can be used for predictive purposes.

134

Coefficient / Constant Value for Test Contractor NPV

Equation Terms

Government NPV

Equation Terms

Sum 151.24 1997.93

Constant N/A -1179.99 -78.29088

Royalty 9% (0.09) -98.80 147.15

CRC 65% (0.65) 197.63 -207.47

CPS 50% (0.5) 499.40 -802.25

Tax Rate 35% (0.35) -306.45 476.20

Uplift Rate 40% (0.4) 87.77 -120.55

Contractor Discount Rate 9% (0.09) -1218.18 N/A

Government Discount Rate 7% (0.07) N/A -3183.46

Average Price 155 1154.21 2939.84

Average Generation 4600 1133.67 2905.32

Table 14: IV values for testing regression equations

Regression Discussion

The results indicate that while using PSA-like terms for the OWF analysed, and

for a 1% change in rate, Discount Rate has the largest effect on the NPV for both

parties. Royalty Rate, Tax Rate, CPS, and Average Price have the next largest effect.

The smallest effects are from CRC and Average Generation. However, since absolute

values of the discount rates vary by less than 4%, and the fiscal terms’ absolute values

vary by up to 65% in the case of CRC, the reader should be cautioned against placing

too much importance on the discount rate.

Secondarily, the discount rate is not a negotiated term in the contract - it is an

internal factor selected by the investor. Comparing the effect of the terms negotiated

135

by the counterparties – namely Royalty, CRC, CPS, and Uplift Rate (and potentially

Tax, though that is often legislated) is more relevant and can make a deal to lease

seabed for an OWF more, or less, attractive to each party.

An interesting discovery from this regression analysis is that for a 1% increase in

certain fiscal levers, the result is not the same for both parties. The two parties’ NPVs

are not equally sensitive to a change in the value of an IV. This is an important factor

to consider when negotiating the PSA. Ceteris paribus, for a 1% increase in Royalty

Rate, Contractor NPV decreases by US$12.52MM, but Government NPV increases by

US$17.61MM. Thus, a negotiated change in Royalty rate is of more value to the

Government than it is a detriment to the Contractor. The Contractor could compromise

by decreasing the Royalty Rate, sacrificing US$12.52MM but get a concession from

the Government of greater value in return, as the Government views that sacrifice as

worth US$17.61MM.

Ceteris paribus, a 1% increase in the CRC has a similar magnitude of beneficial

effect to the Contractor as it has a detrimental effect on the Government. It adds

US$2.89MM to Contractor NPV and removes US$3.24MM from Government NPV.

The two parties should view a percentage change in the term similarly.

Ceteris paribus, a 1% increase in Uplift Rate increases Contractor NPV by

US$2.09MM and decreases Government NPV by US$3.05MM – again the parties

may view this similarly.

Ceteris paribus, a 1% increase in the CPS has a less detrimental effect on the

Contractor (loss of US$10.24MM from NPV) than it does a benefit to the Government

136

(addition of US$15.96MM NPV), so again, a crafty Contractor would sacrifice a

percentage point in CPS to gain something of more value.

Finally, ceteris paribus, a 1% increase in the Tax Rate decreases Contractor NPV

by US$10.23MM and raises Government NPV by US$15.80MM, so if the Tax Rate is

negotiable, the Contractor stands to lose less by a Tax Rate rise than the Government

gains.

Looked at in combination, a Contractor prepared to accept a 1.2% rise in Royalty

Rate (losing US$15.02MM) may wish to negotiate a 1% increase in all three of CPS,

Uplift Rate, and CRC (which results in a combined gain of US$15.22MM). This may

be acceptable to the Government which gains US$21.13MM on the new Royalty Rate

and loses US$22.26MM on the other terms.

It is also interesting to note that in the previous example, a gain of approximately

US$15MM to the Contractor represents 8.3% of their total NPV (US$182MM; mean

NPV the dataset Results_20230901) whereas the equivalent US$22MM for the

Government represents only 1.1% of their total NPV of US$1980MM (mean NPV

from same dataset). This means that the rational Contractor feels more sensitive to

changes in the negotiated rates than the Government.

The other non-contract terms affecting the NPVs are Power Generation and Price.

Ceteris paribus, a 1% increase in Average Power Generation increases Contractor

NPV by US$0.24MM and increases Government NPV by US$0.61MM. It is intuitive

that as more power is produced for sale, both parties benefit financially. It is also

natural that the Government should accrue a larger benefit since they take over 90%

of the total NPV (from Results_20230901; total NPV is US$2161MM, of which

137

US$1980MM flows to the Government). Whilst it is important to know that higher

power generation increases cash flow, it is a factor that affects the site placement and

engineering details of the wind farm, not the contract negotiations.

Similarly, Average Price is an important factor in the NPVs, but also not

controlled by the parties to the PSA-like contract. Ceteris paribus, for a 1% increase in

Average Price, Contractor NPV increases by US$7.26MM, and Government NPV

increases by US$18.33MM. Again, the Government has more to gain from a price

increase in absolute terms but is less sensitive to a price change. The 1% rise in price

increases the Contractor NPV by 3.8% (from US$182MM to US$189MM, using the

mean of Results_20230901) but increases the Government NPV by only 0.9% (from

US$1980MM to US$1998MM, using the mean of Results_20230901). The price for

which the power is sold could be tied to market rates which float daily or could be

fixed at the time of OWF construction through PPAs. These issues should be carefully

considered but like power generation, are not part of the negotiations between

Contractor and Government for the leasing of the seabed.

A final note follows on the discount rates and their effect on the total NPV of the

project. Since the Government Discount Rate is lower than the Contractor’s, the

maximum project NPV can be attained if the entire cash flow is allocated to the

Government. The cash flow would be discounted at a rate in the range of 5.04% to

8.88% rather than the Contractor’s range of 7.04% to 10.88%. A quick comparison,

running 10,000 trials with the following changes revealed the mean NPV results in

Table 15:

138

Case Total

NPV

Government

NPV

Contractor

NPV

1: Independent variable distributions as outlined

in above multiple regression

2,333 2,084 249

2: IVs weighted towards Government; CRC =

0% and mean CPS = 0.001%

2,851 4,309 -1,458

3: IVs weighted towards Contractor; CRC =

100% and mean CPS = 99.9%

2,233 1,446 789

Table 15: Cases illustrating the change in total NPV when fiscal terms are weighted towards either

party

This illustrates the potential for a high total NPV when the lower (Government)

Discount Rate is applied to it (case 2 is higher than case 1 or 3). However, this is a

hypothetical observation, as the Contractor will not enter the project without being

able to demonstrate a positive NPV. Should the Government and Contractor Discount

Rates be equal, this phenomenon would not be observed and the total NPV would not

vary with the contract terms being biased towards one party or the other.

Within the ranges of fiscal terms in this research, and for the OWF analysed,

PSA-like terms can be used as a leasing instrument. It may also be interesting in the

future to perform the same analysis for an oil or gas field, and the conclusions may

have utility for Contractors or Governments negotiating for O&G leases.

Optimal Contract Terms

Regression would encourage the reader to believe there is a symmetrical rise or

fall in NPV for a +/- 1% change in a given IV. However, for Royalty Rate, CRC,

CPS, and Uplift Rate this does not hold true as shown in the following analysis.

139

140

Initially each fiscal term is examined at 1% increments over the whole range used

in this research, to understand its behaviour.

Fiscal Terms in Depth

Royalty Rate

A change in Royalty Rate over the whole range of rates used in the research

(0% to 30%) yields the change in total NPV in Figure 11.

Figure 11: Change in Total NPV for a 1% change in Royalty Rate

Total NPV increases as Royalty Rate rises as more cash flows to the Government,

the party with the lower discount rate. As can been seen, the change in total NPV stays

constant within a certain array of rate changes, then stair-steps to a higher NPV delta

with every few Royalty rate increases. This occurs because the residual revenue after

the deduction of Royalty flows to Cost Recovery. The higher the Royalty, the smaller

the Cost Recovery fund. The Contractor then takes more years to recover its costs.

Every so often, the year in which the cost recovery of capital is completed slips by one

year. As Royalty Rate rises, it can force the final year of Cost Recovery to slip one

Change in Total NPV for each 1% change in Royalty Rate from

0% to 30%

5.5

5.4

5.3

5.2

5.1

4.9

Change in Total NPV US$MM

10%

11%

12%

13%

14%

15%

16%

17%

18%

19%

20%

21%

22%

23%

24%

25%

26%

27%

28%

29%

30%

141

year later, or as the rate decreases, the final Cost Recovery may accelerate to one year

earlier. As a portion of recovered capital becomes cash inflow to the Contractor a year

earlier, it is discounted one year fewer, and adds to NPV. Conversely, if the final year

of Cost Recovery becomes later, that cash inflow to the Contractor is discounted by

one more year, and total NPV decreases. At low Royalty rates, when more funds are

available for Cost Recovery, this slippage occurs less frequently. In this dataset,

between 4% and 10% Royalty Rate (seven different rates), the final year of Cost

Recovery remains constant. However, between 25% and 30% (only five different

rates), the year slips three times as the higher Royalty has a more regressive effect on

the project economics.

Cost Recovery Ceiling

A change in CRC over the whole range of rates used in the research (0% to

100%) yields the change in total NPV in Figure 12.

Figure 12: Change in Total NPV for a 1% change in CRC

Change in Total NPV for each 1% change in CRC from 50% to 100%

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 100

-0.5

-1

-1.5

-2

-2.5

-3

-3.5

-4

-4.5

-5

Change in Total NPV US$MM

142

Total NPV is highest at 0% CRC, as no funds are made available to the

Contractor for Cost Recovery, and they are simply divided 50.85%: 49.15% in the

Profit Split phase. Both the Royalty, and 49.15% of the shared profit are subject to the

Government’s lower discount rate. However, as these projects are uneconomic for the

Contractor they would not proceed, netting both parties nothing. As the CRC rises

from 0%, funds are diverted to the Contractor, and it recovers more costs. It discounts

this cash flow at a higher rate than the Government and total NPV drops. At low

CRCs, the decrease in total NPV is around US$4MM for every 1% rise in CRC, as the

money is used instantly by the Contractor. In this dataset, at 22% CRC, the project

becomes slightly economic, but the whole Cost Recovery fund is still used. At 33%

CRC (the start of the break in curve on the chart) the project creates yet more value

for the Contractor, but there are still some unrecovered costs in the final year. Above

34% the unrecovered costs quickly diminish, and the excess (unrequired) Cost

Recovery funds roll over to Profit Split where they are shared 50.85:49.15. In

profitable projects the Contractor’s Profit Split is then subject to Income Tax, netting

the Contractor less than 50.85% and the Government the balance. The change in total

NPV falls from

~US$4MM to less than US$0.25MM as the funds are diverted through Profit Split and

Tax to the Government, who apply the lower discount rate.

Contractor Profit Split

A change in CPS over the whole range of rates used in the research (20% to

80%) yields the change in total NPV in Figure 13. As CPS rises, more money is

diverted to the Contractor, using the higher discount rate, and total NPV drops.

143

Figure 13: Change in Total NPV for a 1% change in CPS

The change in total NPV is very similar over the whole range – in fact, identical

from 20% to 40% at US$-5.38409MM for each change in CPS. At 41% CPS, the

change in total NPV drops to US$-5.38286MM per 1% change, and from 42% to

80% the total NPV delta drops slightly again, to US$-5.38279MM for each 1%

change in CPS. This level of detail would not be the focus of a negotiation. Since the

Profit Split is calculated after the Cost Recovery mechanism has been completed, the

stair-step effect seen in Royalty and Cost Recovery is not seen. However, the

depreciation for tax purposes takes place after the Profit Split is calculated. In this

dataset between 40% and 41% CPS, the last year in which a tax loss carry-forward

(LCF) is taken moves from 2038 to 2037. Once the LCFs are taken, the remaining

cash flow is taxed and flows to the lower discount rate party, the Government. Hence

the smaller drop in total NPV after the LCFs are completed. The total NPV itself falls

Change in Total NPV for each 1% change in Contractor Profit Split from 20% to

80%

-5.382

-5.3825

-5.383

-5.3835

-5.384

-5.3845

21%

23%

25%

27%

29%

31%

33%

35%

37%

39%

41%

43%

45%

47%

49%

51%

53%

55%

57%

59%

61%

63%

65%

67%

69%

71%

73%

75%

77%

79%

Change in Total NPV US$MM

144

as the CPS rises

145

and the Contractor gets a larger portion of the profit. However, the magnitude of the

fall in total NPV decreases as some of the profit is diverted back to the Government in

Income Tax. It is not apparent what drives the small decrease in the drop in total NPV

between CPSs of 41% and 42%. Since the difference is less than US$70 on a multi-

billion-dollar project the researcher considers uncovering the explanation an unworthy

rabbit-hole.

Uplift Rate

A change in Uplift Rate over the whole range of rates used in the research (0%

to 50%) yields the change in total NPV in Figure 14 14. As the Uplift Rate rises,

more money is diverted to the Contractor, using the higher discount rate, and the

total NPV drops.

Figure 14: Change in Total NPV for a 1% change in Uplift Rate

For Uplift Rate, a stair-step pattern can be seen as for Royalty. As the Uplift Rate

rises, the Contractor recovers more cost, and the recovery process takes longer. At

certain Uplift Rates, Cost Recovery shifts out one year, and those recovered costs are

Change in Total NPV for each 1% change in Uplift Rate from 0%

to 50%

-1.06

-1.08

-1.1

-1.12

-1.14

-1.16

-1.18

-1.2

-1.22

-1.24

Change in Total NPV US$MM

146

subject to more discounting. Therefore, the change in total NPV becomes smaller.

The stair-steps are evenly spread, unlike they were for Royalty. An increase in

Royalty limits the Contractor's ability to recover costs whereas an increase in Uplift

does not – it allows the Contractor to recover ever more cost without limitation in a

profitable project such as this – hence evenly spread stair-steps.

Income Tax Rate

A change in Tax Rate over the whole range of rates used in the research (0% to

50%) yields the change in total NPV in Figure 15. As Tax Rate rises, more money

is diverted to the Government, using the lower discount rate, and total NPV

increases.

Figure 15: Change in Total NPV for a 1% change in Tax Rate

For each change in Tax Rate, the change in Total NPV is the same. Tax is the

last fiscal division to be made in the contract and the cash flow is simply diverted

from one party to the other. Payment of extra tax does not affect any cost recovery

or loss carry forward mechanisms, therefore stair-steps are not seen.

Change in Total NPV for each 1% change in Tax from 0% to 50%

6.00

5.00

4.00

3.00

2.00

1.00

0.00

2% 4% 6% 8% 10%12%14%16%18%20%22%24%26%28%30%32%34%36%38%40%42%44%46%48%50%

Change in Total NPV US$MM

147

For the OWF in this research, each IV behaves differently and does not have

the same effect on the DVs across the whole IV value range. Parties negotiating

PSA-like terms for an OWF should be aware of the idiosyncrasies of each fiscal

term in the range of values he/she is negotiating.

Optimal Terms in a Small Range

Now that the behaviour of the terms has been studied over the same large range

used for the multiple regression, smaller ranges are selected as if in active

negotiation.

Royalty Rate

The change in Contractor, Government, and total NPVs as the Royalty Rate

rises in 2% increments from 8% to 14% can be seen in Figure 16 16.

Figure 16: Change in NPV under the four values ascribed to the IV

The total NPV increases as the Royalty rate rises as more cash flow, in the

form of Royalty, is taken by the Government. The Government has a lower discount

rate than the Contractor, so the extra royalty it receives is discounted less and the

148

total NPV goes up. Contractor NPV decreases as the Royalty Rate rises, as the

Contractor pays more royalty. Conversely, Government NPV increases as more

cash flows to it. Each incremental increase in Royalty Rate does not result in an

equal step in NPV. According to the regression analysis performed for RQ2, we

would expect Contractor NPV to fall US$25.05MM with any 2% increase in

Royalty Rate. Here we see that between 8% and 10%, NPV falls US$24.22MM but

between 12% and 14% NPV falls US$24.92MM. This is asymmetrical – the

Contractor has more to lose from a shift in Royalty from 12% to 14% than from a

2% rise from a lower number. The same asymmetry can be applied to the

Government’s NPV. This asymmetry is expected as it was found in the section

“Fiscal Terms in Depth” that the change in total NPV due to a change in the

Royalty Rate differs depending on the starting rate. As Royalty is calculated before

Cost Recovery happens, the Royalty amount affects the funds available for Cost

Recovery. Therefore, a change in Royalty rate affects the whole Cost Recovery

mechanism and may affect the final year in which capital costs are recovered.

In summary, within the range of 8% to 14%, a 2% rise in Royalty Rate

decreases Contractor NPV by US$24.2MM to US$24.9MM and raises Government

NPV by US$34.5MM to US$35.2MM. Either party will want to consider not only

the change in Royalty Rate, but the starting rate from which that change is made.

Cost Recovery Ceiling

In the range of CRC of 66% to 72%, a change in the Ceiling affects the total

NPV little as can be seen in Figure 17Figure 17.

149

Figure 17: Change in NPV under the four values ascribed to the IV

In this range, the Contractor is able to recover all its capital costs within the

first 10 years of power generation, such that the extra ceiling does not provide much

benefit – from US$4.50MM to US$4.77MM in increased NPV. A small amount of

capital cost can be recovered in an earlier year, thereby gaining some time value of

money. Were the unrecovered costs much higher, the distinction would be more

important – should the CRC move from 0% to 2%, the Contractor NPV increases

US$20.4MM as the Contractor can suddenly recover some costs where it could

recover none before. Contrast this with the regression analysis performed for RQ2,

in which we would expect Contractor NPV to rise US$5.78MM with any 2%

increase in CRC.

Cash earmarked under the CRC but not required to cover costs flows to Profit

Split, where it is split 50.85:49.15 Contractor: Government in this scenario.

Because there is little shift of funds from one party to the other between 66% and

72% CRC, there is only a minor change in the total NPV, as little cash flow is

150

moving from the party with the lower discount rate to the one with the higher

discount rate.

As with Royalty, the change in NPV is not by the same amount for each step in

CRC. For the Government, a rise in CRC from 66% to 68% yields a loss of NPV of

US$5.22MM, but that same magnitude of rise from 70% to 72% yields only a

US$4.89MM loss. The reason is the same as for Royalty – as the CRC increases,

the final year of recovery of capital costs accelerates, and each time it moves

forward one year, the total project NPV increases as less discounting is applied.

In summary, within the range of 66% to 72%, a 2% rise in CRC increases

Contractor NPV by US$4.5MM to US$4.7MM and decreases Government NPV by

US$4.9MM to US$5.2MM. Again, either party will want to consider not only the

change in CRC, but the starting rate from which that change is made.

Contractor Profit Split

Conversely, changes in NPV due to a change in the CPS are very close to

symmetrical. In the range of 48% to 54% CPS, it is necessary to look to the twelfth

decimal place to see a difference in the change in total NPV. As expected, for a rise

in CPS, Contractor NPV increases and Government NPV decreases. As cash is

flowing from the lower discount rate party to the higher discount rate party, total

NPV falls, as can be seen in Figure 18.

151

Figure 18: Change in NPV under the four values ascribed to the IV

This symmetry is to be expected in the range 48% to 54% as discovered in the

section “Fiscal Terms in Depth” as the profit split is a year-by-year calculation and

not affected by the loss carry forward mechanism that affects Cost Recovery and

Uplift.

In summary, within the range of 48% to 54%, a 2% rise in CPS increases

Contractor NPV by US$19.0MM and decreases Government NPV by US$29.7MM.

This time the change in CPS is of importance in negotiations, but the effect does not

differ according to the starting rate.

Uplift Rate

In the range 38% to 44 %, changes in Uplift Rate do not affect NPV

symmetrically as can be seen in Figure 19. A rise in Uplift Rate causes Contractor

NPV to increase as it can take more cash flow from the project through the cost

recovery mechanism. Government NPV falls, but by more as cash flow shifts from

152

the low discount rate party to the higher one. This results in a loss of total NPV as

the Uplift Rate increases.

Figure 19: Change in NPV under the four values ascribed to the IV

The changes are not symmetrical, as seen for Royalty and CRC. The reason is

the same – as the Uplift Rate increases, the final year of recovery of capital costs

may fall later. Each time it moves out one year, the total project NPV decreases as

one year more discounting is applied. The symmetry seen in Figure 19 between

38%, 40%, and 42% is merely because those rates belong to one flat section of the

stair-step chart seen earlier in “Fiscal Terms in Depth”. A stair-step is seen between

42% and 44% explaining the differences in the grey bars of Figure 19.

In summary, within the range of 38% to 44%, a 2% rise in Uplift Rate

increases Contractor NPV by US$4.0MM to US$4.2MM and decreases Government

NPV by US$6.2MM to US$6.5MM. As with Royalty and CRC, it is important to

know not only the change in NPV associated with a change in Uplift Rate but also

the starting point, as that affects the delta.

153

Income Tax Rate

As for CPS, changes in NPV due to a change in the Tax Rate are very close to

symmetrical. In the range of 28% to 34% Tax Rate, the reader must look again to

twelve decimal places in the change in total NPV to see a difference. As expected,

for a rise in Tax Rate, Contractor NPV decreases and Government NPV increases.

As cash is flowing from the higher discount rate party to the lower discount rate

party, total NPV rises, as can be seen in Figure 20.

Figure 20: Change in NPV under the four values ascribed to the IV

This symmetry is to be expected in the range 48% to 54%, as discovered in the

section “Fiscal Terms in Depth”. Tax is a year-by-year calculation and not affected

by the loss carry-forward mechanism that affects Cost Recovery and Uplift. Much

as the change in NPV is flat across all Tax Rates, it is symmetrical in this smaller

range. In summary, within the range of 28% to 34%, a 2% rise in Tax Rate

decreases Contractor NPV by US$19.3MM and increases Government NPV by

US$30.1MM. As for CPS, negotiators must be aware of the change in NPV due to a

154

change in Tax, but the starting point from which that change ensues does not affect

the gain or loss of NPV.

Comparison to Regression Coefficients

A comparison was made, for the sake of completeness, between the results of this

analysis and the regression coefficients (Table 16).

Fiscal Term (IV) Change in NPV for 1%

Change in IV from Multiple

Regression Contractor /

Government (in US$MM)

Average of Change in NPV for

1% Change in IV in the ranges

analysed above

Contractor / Gov’t (in US$MM)

Royalty -12.52 / 17.61 -12.30 / 17.45

CRC 2.89 / -3.24 2.32 / -2.53

CPS 10.24 / -15.97 9.48 / -14.86

Tax Rate -10.23 / 15.80 -9.65 / -15.03

Uplift Rate 2.09 / -3.05 2.08 / -3.21

Table 16: Comparison of multiple regression coefficients and the results of the scenario analysis of each

fiscal term

The multiple regression coefficients are similar, but not identical, to the results of

the scenario analysis of a small range of each IV. For simplicity, or at the start of a

negotiation, multiple regression could be employed to gain a sense of the impact of

each fiscal term. It would provide an approximation and allow the negotiator to make

broad statements without running the cash flow model each time. Later, when

precision is required during the final stages of negotiation, scenario analysis is a

better choice of tool to assist detailed understanding of the effect of changes in

contract terms.

155

Use in Negotiations

The following example shows the scenario analysis at work in an illustrative

negotiation. Table 17 shows the mean change in each NPV for a 2-percentage point

change in each fiscal term, as analysed earlier in “Optimal Terms in a Small

Range”. To reiterate, a change in Royalty Rate, CPS, and Tax Rate have the biggest

fiscal term impact on the NPVs.

Fiscal Term / IV Mean Change in NPV in US$MM

Total Contractor Government

Royalty Rate 10.29 -24.60 34.89

CRC -0.43 4.64 -5.06

CPS -10.77 18.95 -29.72

Uplift Rate -2.27 4.15 -6.41

Tax Rate 10.77 -19.29 30.06

Table 17: Mean change in NPV for a 2% change in IV value within the ranges discussed above

A suggested change in a term during negotiations that harmed one party could

be recovered by changes to other terms. For example, a proposed 2% increase in

Royalty Rate sees the Contractor forfeit US$24.60MM. This could be recovered in

the following ways (Table 18) or an unlimited combination of fractions of more

than one term:

156

Fiscal Term Required % change Resulting gain in NPV

CRC (option A) +10.60% US$24.60MM

CPS (option B) +2.60% US$24.60MM

Uplift Rate (option C) +6.36% US$24.60MM

Tax Rate (option D) -2.55% US$24.60MM

Table 18: Required changes in three IVs to equal a +2% change in Royalty Rate

A sensible approach to negotiations might be to request a 2% increase in CPS

(to match the 2% rise in Royalty Rate) in conjunction with a 2.72% increase in

Uplift Rate. This would also recover US$18.95MM + US$5.65MM =

US$24.60MM (option E).

The 2% rise in Royalty Rate affects the Contractor NPV differently from the

Government’s – its NPV increases by US$34.89MM (Table 17). Table 19 shows the

effect on the Government NPV of the five above-mentioned counteroffers that the

Contractor might make.

Fiscal Term(s) Change in Rate Change in Government NPV

CRC (option A) +10.60% US$-26.80MM

CPS (option B) +2.60% US$-38.64MM

Uplift Rate (option C) +6.36% US$-20.38MM

Tax Rate (option D) +2.55% US$-39.02MM

Combination (option E) +2% CPS, +2.72% Uplift US$-38.44MM

Table 19: The effect of the four negotiations options on Government NPV

Clearly, options B, D, and E are the worst as they require that the Government

give up more value than it has gained in requesting the Royalty Rate increase.

157

Options A and C, however, allow the Contractor to retain their former NPV whilst

allowing the Government to increase theirs – their loss is smaller than the gain from

the Royalty Rate rise. The Contractor has room to manoeuvre – it could request an

increase in CRC above 10.6% or an increase in Uplift Rate above 6.36% and

increase its NPV whilst also allowing the Government to increase its NPV. This is

the essence of a “win-win” negotiation and the formulation of optimal contract

terms.

The reader with a keen interest in economics will have noticed that the

Government’s desire for a gross Royalty, and its use of a lower discount rate than a

commercial enterprise, are inconsistent. The Royalty gives the Government cash

quickly – from the outset of the OWF coming onstream and before the Contractor

recovers costs. It is perhaps not surprising that a government seeks cash in any

given fiscal period, to balance the national budget and provide for their country.

Cash inflow may be an even higher priority for developing countries, with a weaker

economic base, lower GDP per capita, and less/little access to capital markets. A

desire for re-election may also drive the preference for being seen to gain early cash

flow from lucrative projects harvesting national resources. Yet commonly

governments use low discount rates to evaluate O&G projects into which they are

investing no capital, and in which IRR is therefore meaningless (researcher’s

experience). This discrepancy could be the topic of a dissertation in itself.

Stochastic Analysis for Negotiation

In a real negotiation, the contracting parties would wish to run the cash flow

model with the fiscal terms set to the rates under discussion at the time, and let the

158

other IVs vary within the full distribution considered applicable to the OWF in

question. Here, the fiscal term IVs were set to their average rates to simulate an active

negotiation, and the other five IVs varied according to their distributions. The model

was run 10,000 times (dataset: Results_20231027) and an NPV distribution was found

for each party (Figures 21 and 22).

Figure 21: Distribution of Contractor NPVs

159

Figure 22: Distribution of Government NPVs

The mean of the Contractor NPVs is US$305MM with a standard deviation of

US$144MM and a range of US$-297MM to US$735MM. The mean of the

Government NPVs is US$2,057MM with a standard deviation of US$320MM and a

range of US$912MM to US$3,085MM. These statistics, together with myriad other

data points could be used for decision making. For example, the Contractor may wish

to calculate the proportion of negative NPV results to gauge risk. The NPV ranges

alone offer more sophistication than the single-point NPVs often used in the O&G

industry.

NPV Sensitivity to Project Characteristics

The “base case” is defined as the case where all IVs take their mean value.

Sensitivities to a 20% increase or decrease in the IVs Price, Cost, and Generation are

shown in Figure 23. At the centre of the chart, the US$0 NPV line represents the base

Change in NPV from the Base Case

High CaseLow Case

250200150100500-50-100-150-200-250-300

C (+20%/-20%)

P (-20%/+20)

G (-20%/+20)

The Effect on Total NPV of a 20% change in Generation, Price and

Cost

160

case. Ceteris paribus, a 20% decrease in Generation or Price has a negative effect on

the NPV, shown in red, and known as the “low case”. A 20% increase in Cost (in red,

the “low case”) also harms NPV. 20% increases in Generation and Price, and a 20%

decrease in Cost (“high cases”) have a positive effect on NPV and are shown in green.

Figure 23: The effect of a 20% increase and 20% decrease in Price, Cost and Generation on Total NPV

For this dataset, the change in Generation and Price have equal effects. This is to be

expected as both affect gross revenue, calculated as Generation multiplied by Price. A

20% change in one is equivalent to a 20% change in the other. The total NPV is less

sensitive to a change in Cost. This is interesting to note for the Contractor who may

wish to pay more attention to Price and Generation than Cost when considering risk in

the low cases. High cases are a happy surprise and typically less scrutinised in running

project economics before a decision (researcher’s experience). Generation is partially

under the Contractors’ control in that it designs the maximum capacity of the wind

farm. However, how much wind resource is present at any given time is not under the

161

Contractor’s control. Price can either be completely under the Contractor's control or

162

not at all. If it lets the sales price float with the market price it has no control. If it

enters a PPA or other fixed-price sales contract, the Contractor can have certainty over

the Price and eliminate the risk (and upside) illustrated in Figure 23. To a degree, Cost

can be controlled and known by the Contractor. Contracts will be signed for all

elements of the building and maintenance. Delays and associated cost overruns could

affect total cost but careful management by the Contractor can minimise those.

Future work could re-introduce the stochastic element to this analysis by using a

distribution for the change to the IVs instead of a flat +/-20%. The Contractor and

Government would then know the likelihood of certain events happening.

Another interesting data point is the breakeven mark for each of the IVs. Table 20

shows the percentage change and resultant IV value at which the total NPV becomes

US$0 for each IV, ceteris paribus.

Independent Variable % of mean value at which

total NPV reaches US$0

Resultant value of the IV

Cost 231.14% US$11,365MM

Price 43.27% US$67.84/MWh

Generation 43.27% 2,027.6 GWh/year

Table 20: Percentage changes and breakeven values for Cost, Price, and Generation when total NPV is

zero

As can be seen from Table 20, Cost must more than double to render the project

valueless. Price and Generation (again equivalent for the reason stated above) can fall

to below half their mean values before total project NPV reaches US$0. However, in

all three cases, the Contractor NPV reaches US$0 before the total project NPV

reaches

163

that point. Table 21 shows the split of NPVs between Contractor and Government for

each of the scenarios in Table 20.

Scenario / Independent

Variable

Total NPV Contractor NPV Government NPV

Cost US$0 US$-1,194.89MM US$1,194.89MM

Price US$0 US$-516.97MM US$516.97MM

Generation US$0 US$-516.97MM US$516.97MM

Table 21: Split of total NPV by party when total NPV is zero under the Cost, Price, and Generation

scenarios

Again, the results for Price and Generation are equal due to gross revenue being

the product of the two. In these two scenarios, the split is less uneven between the two

parties than in the Cost scenario. This is not surprising as revenue is split between the

parties by the various fiscal term mechanisms. Cost, however, is entirely borne by the

Contractor, so a 231% increase damages the Contractor’s NPV more greatly.

The Government still has a positive NPV in all cases and whilst it may dislike

the lower NPV, it would rationally still participate in the projects. From the

Contractor’s perspective, however, it is important to know the breakeven points of the

three IVs that cause Contractor NPV to reach US$0. Table 22 shows these results.

164

Scenario /

Independent

Variable

% of mean value at

which Contractor

NPV reaches US$0

Resultant total /

Government NPV

Resultant value of

the IV

Cost 129.37% US$1,663.08MM US$6,361MM

Price 77.30% US$1,285.56MM US$121.21/MWh

Generation 77.30% US$1,285.56MM 3,622.7 GWh/year

Table 22: Percentage changes and breakeven values for Cost, Price and Generation, and

Total/Government NPV, when Contractor NPV is zero

The project can only sustain smaller deleterious changes before Contractor NPV

reaches US$0 and the energy company would decide not to participate in the wind

farm. A 30% Cost overrun or a 23% decrease in Price or Generation renders the

project valueless from the Contractor’s perspective. Since the Government has a

positive NPV in all three cases, there is room to negotiate less regressive fiscal terms

which would better distribute the modest total NPV between the two parties.

Removing Royalty and increasing the Uplift Rate would assist in this goal. For

example, adjusting the “low” Generation scenario, giving it 0% Royalty and 80%

Uplift (vs. the means of 8.5% and 40% respectively) would split the NPV and award

US$144MM to the Contractor and US$1,068MM to the Government. The rational

Contractor would again be willing to proceed.

In summary, this OWF is more sensitive to changes in Cost than Price or

Generation. Marginal projects, where any of those three IVs are close to the breakeven

point, can be improved by fine-tuning the fiscal terms. Where project outcomes are

not known before building the OWF, the contract terms should be tested for both

165

marginal

166

and highly profitable projects. Progressive fiscal terms, such as low Royalties and the

Profit Split found in PSAs, can ensure marginal projects create value for the

Contractor and that most of the upside goes to the Government in a success case.

Views of PSAs for O&G and OWE

Initial Salient Points

Word frequency within the interviews was analysed and the results and

discussion can be found in Appendix 20. In summary, the salient concepts were

similar regardless of whether the participant was speaking of O&G or OWE

indicating the similarity between the two applications. Next, during the axial coding

phase, the first response that each participant gave to the questions was collected as

that may indicate their most important thought. The results can be found in Table

23.

Question Concept Number of First

Mentions

PSA Advantages in O&G Local content,

infrastructure provision,

jobs, and training

Fiscal Terms - Profit Split,

Cost Recovery

Host ownership of

resources

167

Bring expertise and

resources together

Risk mitigation 2

Contract Stability 1

PSA Disadvantages in O&G Limited upside for

contractor

Contract instability 1

“Gold plating” – gaming

the contract

PSA Applicability to OWE Downside protection, cost

recovery for Contractor

Bring expertise and

resources together

Host participation and

knowledge transfer

Table 23: Number of times a concept appeared as the first response to a question

These first-raised themes hint at the details of the interviews to be analysed in

the sections below. These include the importance of such concepts as the workings

of the financial and non-financial terms of the contract, the protection it offers, and

the fact it brings together essential and lucrative natural resources with the financial

capacity and skills needed to harvest them.

168

Advantages of PSAs in Oil & Gas

Eight clear themes - advantages of PSAs in O&G - emerged, with two others

each mentioned only once. The first of these lesser themes is the DMO, whereby the

host country is allowed by the contract to buy a typically small portion of the oil

and/or gas production at a discounted price for domestic consumption. The second

is the fact that the Contractor controls the operations, can control costs to the best of

their ability, and apply their preferred safety standards, which are often higher than

local regulations require.

The eight primary advantages were mentioned with the frequencies shown in

Table 24 below.

Theme Number of references

Non-financial benefits Contractor provides 19

Host low-risk, low-cost participation 16

Flexibility and progressivity of fiscal terms 18

Profit Split 14

Cost Recovery 13

PSA as a contract 11

Trade Expertise & Knowledge Transfer for Resource 11

Resource entitlement and retention 8

Table 24: The eight primary themes within advantages of PSAs for O&G, and the frequency with which

they appear

Non-financial benefits provided by the Contractor

According to the participants, non-financial benefits of the PSA are an important,

and unusual (compared to a CA) part of the structure. One participant summarises

169

nicely, describing them as “non-traditional monetary gains that a PSC typically

[includes] like local content… which includes jobs, building infrastructure, education,

schools.”

These benefits include local or national content, where local people and

fabrication facilities must be employed in the construction and operation of the O&G

field. A caveat raised was the inability of an underdeveloped local industry to supply

the goods and services required by the project.

Another benefit is the training and development of local O&G professionals

which allows them to climb the skills curve and eventually perform work as a

Contractor themselves. One activity that helps fulfil this is the secondment of NOC

employees directly into the project, rather than simply as suppliers for local content.

This is accomplished through Government participation in the project, which results

in the training of locals. Another benefit is direct grants to educational establishments

or education budgets to train the next generation of O&G workers in the secondary

and tertiary education systems.

Requirements for local infrastructure benefitting the citizenry such as roads,

harbours, schools, clinics, and water wells may be included. These may be key to the

construction project (such as a harbour or road) or purely for the betterment of the

populace such as schools, clinics, and wells.

Participant comments about these initiatives were “it’s fantastic”, “it’s fair” and

“makes sense”. It was also noted that these benefits are valued more highly by

170

developing countries than developed countries. The latter do not need new

infrastructure, education, or training, and typically use the CA regime, not PSAs.

Host low-risk, low-cost participation

Participants talked about the clause in many PSA contracts that allows the

Government, through the NOC, to take an equity stake in the project. This provides

the learning opportunities - “a training run”- described in the above section. It also

delivers another source of income for the Government – not just through cash flow

from the Royalty, Profit Split, and Tax, but an additional Profit Split from being an

asset owner. NOC participation is carried i.e., the other owners pay their costs, and

those owners recover what they paid from the Government’s Profit Split. This

eliminates the Government’s financial risk and expense in the setting up of the O&G

field, which will be popular with voters. The Government also gains from the time

value of money, as the Contractor makes up-front expenditure on their behalf and the

Government repays it later, from profits.

The intent of the participation in the long run is for a transfer of knowledge to the

NOC to enable it to carry out operations independently in the future, as has been seen

in Brazil, according to one participant. This could be later in the life of the same field,

or it could in replicating the work on a similar new field. Several participants

expressed concern that this process must be done slowly in order that the NOC

becomes fully capable of the intricate and potentially dangerous operations before

taking over. There is typically flexibility in the contract such that if the Government is

not ready to take over, they do not – thereby continuing to enjoy low-risk operations.

171

Another advantage stated is that having an NOC typically makes a country’s

regulatory body stronger as the government has deeper involvement in the O&G

sector. This ties in with Antolin et al’s case study on Brazil and Bolivia.

Cost Recovery

Cost recovery is the unique contract concept that allows the Contractor, who has

made all the expenditures, to recoup their costs from the cash inflow to the project

before any profits are disbursed (Tordo). The amount of cash flow available depends

on the CRC, which can be set to 100%, in which case all the post-Royalty cash flow is

available for cost recovery. However, it is frequently less than 100% to allow some

cash to flow immediately to the profit-sharing pool. Five of the participants

recognised the power of the Cost Recovery mechanism in decreasing the risk for the

Contractor, saying that the “cost recovery… is for the company”, “they [Contractors]

will want that [CRC] to be as high as possible”, and “it provides some assurance to

the investing oil companies that they'll recover these multibillion-dollar investments

typical for an offshore contract”. Another recognised the upside for the Government,

stating a preference for a CRC less than 100%, as it provides early cash flow to the

Government through Profit Sharing. One participant identified that the protection for

the Contractor in Cost Recovery is also good for the Government in that a marginal

project is more likely to go ahead when the Contractor is comfortable. Another

touched on the complexity of Cost Recovery as a downside protection mechanism for

the Contractor, in that the recovery is given in barrels of oil or cubic feet of gas.

Should the commodity price fall, the Contractor will receive more hydrocarbons to

pay back their expenditure than if the oil/gas price was high.

172

Profit Sharing

Six participants identified profit sharing, or profit split (terms used

synonymously) as an important feature of the contract. One expressed that “the well-

constructed PSC [common industry synonym for PSA] tends to redistribute those

returns to the host country… that's the attraction to the PSC structure”. The profit

after Royalty and costs are (re)paid is split by a pre-determined ratio between

Contractor and Government. In contracts in 74 countries such as Angola, Azerbaijan,

Bangladesh, Ecuador, and India, the split is not fixed but moves on a sliding scale

based on IRR, R-Factor (cumulative revenue / cumulative costs), production rate,

revenue, cumulative production, and price (Wood Mackenzie). The common theme

articulated by participants is that the financial upside – the bulk of the profits,

especially when they are high – goes to the Government, to whom the resource

belongs. Two participants commented that energy companies naturally do not like this

feature. Several participants pointed to the poor reputation of O&G companies for

taking advantage of less savvy nations and seeking “super-profits” on O&G projects.

However, one said that this method of profit-sharing acts as a good safeguard of the

nation’s resources, and in turn may allow projects to go forward that the citizens

would not otherwise have welcomed. Several participants cautioned that the profit-

sharing formula must be carefully crafted in order to achieve its intent. Enough of the

revenue must be available to compensate the Contractor through Cost Recovery; then

the remaining revenue must be judiciously split to encourage the Contractor to take

project risks, whilst the upside flows to the resource owner. This upside could occur

in an unexpected high-price environment, when production is higher than expected, or

173

costs are lower than expected. Some participants, despite being O&G company

employees, recognised that this situation is fairer than a CA regime, in which the

upside flows to the Contractor. One participant took the concept a step further, in

saying that a government which knows it is benefiting handsomely in an upside case is

less likely to try and re-negotiate the contract. He pointed to the UK regime, a non-

contractual CA regime, as one in which a windfall profits tax has been introduced

multiple times throughout history. This creates uncertainty for the Contractor at the

outset of a project in not knowing exactly how it will be taxed.

As a final note, one participant did remind us that the profit split received by the

Contractor is then subject to Income Tax and is therefore less than it appears in the

contract. Assuming the company is in a taxable position in the country involved, this

is true. This was also noted in the quantitative analysis forming part of this research.

Flexibility, Progressivity, and Fairness of Fiscal Terms

The primary aspect of progressivity of fiscal terms mentioned by five participants

is that the Royalty Rate can remain low because there is another mechanism through

which the Government will receive cash. In a CA regime, the only opportunities for

the Government to take cash from a producing field is through Royalty or Tax. If a

company has many tax deductions available to them, tax may be low, and Royalty is

the only avenue. The incentive for the Government will be for a high Royalty Rate to

ensure the nation some income. However, since Royalty is calculated from gross

revenue and before costs are taken into account, it is a regressive fiscal term, and can

make marginal projects sub-economic very quickly. A low Royalty Rate, that will

ensure all projects go ahead, will not allow the Government to enjoy in the financial

174

upside of a successful project. The Wood Mackenzie data illustrates this point – the

average Royalty Rate for a CA regime is 11.3% whereas for a PSA regime, it is 8.5%.

The participants point to the fact that a PSA can endure a low Royalty Rate, and

encourage marginal projects, but has a progressive term - the Profit Split – which can

be set at a level that allows the Government to take the bulk of the upsides after Cost

Recovery. One participant mentions the inherent fairness that the Government will get

some cash inflow from day one (from the Royalty) plus the majority of the post-cost

profits through profit-sharing “and that is the way they could sell it [the project] to a

poor population”.

The participants also saw two further ways in which the PSA structure

provides flexibility. With five or more fiscal levers at their disposal, there are

myriad ways in which to split the total project NPV, and a good, or optimal way can

be sought.

The second viewpoint holds that since a PSA is a contract, negotiated for every

opportunity, it can vary much more greatly than a country-wide regime such as would

be used for CA. Thus, the PSA can be fit for purpose for one project or area where a

CA cannot. Connected to that, another view offered was that an unsuccessful PSA

offering (if no Contractors bid, or step forward to negotiate) can be changed and re-

offered with different terms without changing the whole legislative regime, as would

happen for a CA.

The participants frequently alluded to the fact that these mechanisms result in the

PSA being a fair contract. One participant went as far as to say, “I tend to think that

more oil and gas people probably would use the word fairness more with a PSC than

they would with a with a concessionary system [CA]”.

175

Lest the reader get too enthusiastic, two participants reminded us that poorly

structured fiscal terms can also create the opposite situation from that which was

intended, and that care must be taken in setting up the contract. This concept was

also illustrated throughout the quantitative research phase.

PSA as a contract

The advantage that a PSA is a contract, and the stability that offers, arose six

times during the interviews. Not only fiscal stability – the Contractor can calculate the

expected NPV for the project or contract area as the terms are known – but also the

assurance of knowing that access to the acreage is unlikely to get revoked. Of course,

this can happen, but is more likely under a CA, where a government elects to make

sweeping changes. An example was the Venezuelan Government which from 2004

re- nationalised oil fields to which ConocoPhillips, Chevron, ExxonMobil, BP, Statoil

Hydro, Petrobras, and Total had rights under the CA regime (Eljuri & Trevino, 2009).

One interesting point made is that PSAs can replace weak or immature legislation

in a poorly developed country. One set of weak or poorly constructed rules applying to

the whole sector could destroy a nascent industry. A country, perhaps a developing

country, might benefit from negotiating each case separately and learning as it does so.

Two participants pointed to the flipside of contract stability - that an agreement

can be re-negotiated if needed, more easily than lobbying for a change in country-wide

legislation. Towards the end of a PSA, close to the end of the field life, the Contractor

might seek changes to the contract to extend the life of the field. For example, if the

Royalty Rate is such that it has rendered the remaining reserves uneconomic, the

Contractor will want to cease production and decommission the field. This leaves

176

resources in the ground which is a disadvantage to the Government. Renegotiation of

one term in one contract can be accomplished easily enough. Contrast this to

renegotiation of a Royalty in a CA regime which could affect all concession holders –

that would be a much harder task.

Finally, one participant mentioned the convenience of the PSA contract as a

manual for operations, or a checklist. All the rights and obligations of the parties are

contained therein (Johnston), and it can be a convenient guide on how to proceed.

Trade Expertise & Knowledge for Resource

A participant explained, “the reason PSAs exist is because you have one party

that has resources and insufficient expertise, and another party that has sufficient

expertise, but needs more access to resources”. It is true that no party – neither

governmental nor business – would undertake to explore for hydrocarbons without

experience and expertise. It is also the case that globally, government ownership of

resources is almost ubiquitous. Granted, experienced companies can be given access to

resources through CAs or other mechanisms, but in none of those instances is there a

trade of anything other than money for resources. PSAs offer the opportunity to trade

access to resources for not only financial payment but also knowledge transfer – a

“give-get” as another participant described it. The details of this knowledge transfer

are covered in the section on non-financial benefits provided by the Contractor. The

key ideas coming out in this theme were that national human capital gets developed

through education, training, development, and secondment (mentioned seven times in

reference to the trade) and through NOC participation in the projects (mentioned four

times). The story of the rise to great success of Petrobras, the Brazilian NOC, through

177

learning from IOCs was given by one participant, “…argue that Petrobras now is one

of the premier deepwater offshore drillers in the world because of the experience that

it had, and … a stake in the contracts”. Another participant more broadly asserts that a

country without “technological sophistication” benefits from this exchange of

knowledge for access to resources. Yet another participant sees this process as a

stepping-stone to NOCs taking over operations later in a hydrocarbon field’s life and

becoming self-sufficient.

Resource Entitlement and Retention

A smaller theme, mentioned eight times, and one unique to the PSA, is that the

host country retains ownership of the resources – the buried hydrocarbons – until a

portion is turned over to the Contractor as re-payment for costs or as Profit Split. As

one participant astutely put it, “politically it’s a very large advantage for them, they

get a lot of trouble internally and politically for selling off the country's resources”. A

second participant alluded to the same, “…give the impression that you are not just

keeping dollars, but you are actually keeping the country's natural resources”. This

may just be optics, but to a proud developing nation suspicious of “Big Oil” entering

the country and putting money before social considerations, it is a “win” for the

country than cannot be gained through other leasing mechanisms.

Disadvantages of PSAs in Oil & Gas

Fewer themes around the disadvantages emerged, and only one had a similar

amount of coverage as the advantages. The four disadvantages were mentioned with

the frequencies shown in Table 25 below.

178

Theme Number of references

Highly complex contract 15

Limited Contractor upside 8

Abuse or instability of contract 6

Contractor entitlement in kind 3

Table 25: The four primary themes within disadvantages of PSAs for O&G, and the frequency with

which they appear

Highly Complex Contract

This theme is split into three concepts and the most frequently mentioned was

the PSA being “less transparent” and “nebulous”. It is more open to interpretation

than a CA regime and harder to understand, for both the users of the contract and

the shareholders in the O&G companies. It is more difficult to negotiate and

navigate legally should the need arise. This is not surprising as the PSA offers a

partnership between counterparties, unlike a CA. A company, or a government,

party to several PSAs in a country must understand each. In a CA regime, both the

Contractor and Government would only have to be experts in that one regime.

The second angle is that the contract is complex to understand; making it

challenging to administrate the activities, to model financially for decision-making,

and to complete the accounting after the fact. Participants with experience as

accountants and financial modellers in both PSAs and CAs made similar comments.

The final observation is that PSAs create more work for the Contractor and are

tough to administrate and remain in compliance with, as they contain many

requirements to fulfil. Local content, education and training, infrastructure building,

and knowledge transfer all require effort and administration, plus extra work to

179

180

demonstrate compliance with the contract. The job of tracking expenditure for cost

recovery often falls to entire departments of an O&G company if that company

holds multiple contracts. Working alongside inexperienced staff seconded from the

NOC to the project and training them requires time and effort to be done properly.

Limited Contractor Upside

The flipside to the benefit for the host country of taking the bulk of the profit in

a highly lucrative project is that the Contractor’s profits are limited. The company

continues to get a portion of the profits as they rise, but the proportion often

becomes smaller and smaller. Consider the case of [country name redacted1], where

the Contractor takes 51% of the profit at low production rates or price levels but

takes only 10% of highly profitable projects (Wood Mackenzie). One participant

referred to the excess profits as “a buffer” for the Contractor in this context.

However, when describing the profit-sharing mechanism as an advantage to the

Government the profits had been described as “super-profits” and “excessive

returns”. It seems it is a matter of perspective, but perhaps it is not surprising that a

clear advantage for one party is seen as a disadvantage for the other. The

enthusiasm with which the participants spoke of the benefit for the Government

outweighed the slight resignation with which they spoke of the same characteristic

being a disadvantage for the Contractor, suggesting the structure is a good one. A

specific point made by one participant was that capping the upside might have the

effect of making the investor, the Contractor, more risk averse. Since O&G

exploration is an inherently financially risky business, this could lead investors to

181

look to explore in countries with a regime that allows the Contractor to capitalise on

the financial upside as a reward.

The other negative stance taken by the participants on the limiting of the

Contractor upside was the fact that Income Tax is paid out of the Profit Split

portion, if the company is in a taxable position. As previously mentioned, this

means that a nameplate 50% Profit Split can become significantly less if, on

average, 31% tax is levied.

Abuse or Instability of Contract

The participants highlighted a few ways in which the contract can be abused,

starting with the occasional propensity of unethical local officials to require that the

budgets for local improvements be spent on items that only benefit specific

individuals. An example given was a West African official who pushed for a villa in

France to be purchased from money earmarked for local improvements.

The second abuse lies with the Contractor, who may be inclined to “gold

plate” items of expenditure in order to recover more costs. The items would have

specifications that are higher than strictly necessary, but since those higher costs

can be recovered, the Contractor is ambivalent to the wasted expenditure, or can

even use the additional expense to increase project quality and profitability

(Johnston & Johnston).

The sanctity of the contract, which is not always honoured, was the final point

raised. Two participants indicated that PSAs are often in countries where the rule of

law is less developed, so there may be a higher risk than in OECD countries of the

182

Government reneging on or changing the deal, or regulatory agencies or local

businesses not abiding by the contract.

Contractor Entitlement in Kind

This topic was a minor one but was raised by two participants on three

occasions. The O&G company’s financial statements can only reflect the resources

to which they are entitled in a PSA, and that is only a portion of the total resources.

Contrast that to a CA in which they can book all the resources in the O&G field, as

the Government does not retain ownership. One participant commented that this

issue is less important than in the past, but the booking of reserves to indicate the

size of the company is a big annual activity and is viewed critically by shareholders

(Researcher’s experience).

Participants’ Thoughts on Renewable Energy and the Transition

The interviews yielded two primary reasons for the change to renewable

energy, two sets of characteristics of the transition, and two unresolved talking

points, as shown in Table 26.

183

Theme Number of references

Reason #1: Commitment to the Environment 14

Reason #2: Socio-politics 4

Characteristics #1: Financial Outlook of Transition 12

Characteristics #2: Timing & Pace of Transition 11

Talking Points 2

Table 26: The themes emerging from discussion of the transition to renewable energy

Commitment to the Environment

Participants talked of the transition in terms of the “energy mix” and “energy

matrix”, commenting that the transition, as the name implies, is a phasing out of

environmentally damaging energy sources and their replacement with less harmful

ones. One talked of an expansion – in the number of sources, and also an overall

expansion as energy demand grows. Another spoke of the clear necessity for the

transition, given climate science data. One talked of traditional O&G companies

committing to climate goals such as the Paris Accord and carbon neutrality which

seem in conflict with their core business. He said, “…the end game is not the

same … as … being a pure oil and gas company anymore… It's a lot about

sustainability, the environment, SSU, things like that”. Three participants expressed

positive views of OWE as a part of the new energy matrix, including “it’s a

wonderful, wonderful thing” and “offshore wind farms will definitely be a big part

of this mix”.

184

Despite being O&G workers, with over 200 years’ experience between them,

the participants have reasonable knowledge of, and a positive disposition towards,

the transition and OWE.

Socio-politics

Three participants mentioned political angles that drive the transition. The first

involves “telling people what they want to hear” – politicians speaking publicly of

the transition to gain popularity, then policy following the promises that have been

made. The others include energy security, diversity, equity, and affordability. These

topics were discussed earlier in Chapter 2. Renewable energy, including offshore

wind can improve a nation’s security and diversity of supply, can provide cheaper

energy sources, and can provide power where perhaps there was no access before.

Financial Outlook of the Transition

In terms of characteristics of the transition; costs, returns, and subsidies came

up frequently. Three participants cited the fact that most projects need subsidies and

one pointed out that subsidies can serve to obfuscate the true returns of the projects.

Four participants talked about the costs involved in building renewable energy

plants with a negative tone, but interestingly, two of them referred to the high cost

of leasing seabed for OWE projects, which itself implies competition. Willingness

to pay a large access cost implies the buyer sees the opportunity for financial return,

or else they would bid lower, or not bid at all, and pursue other lines of business.

Perhaps these companies see the falling costs of offshore wind, and the pressure for

transition, and are prepared to hold the land and wait until their project is

economically viable.

185

On the inferior returns, one participant said, “the returns are really

questionable, especially versus oil and gas projects” and another stated “it's

difficult to find profitable projects within the renewables space”. She went on to

admit that a lower rate of return is acceptable as the risks are less for renewables

than O&G. This acceptability is due to the renewable resources being demonstrable

ahead of time, whereas O&G is only proven after expensive operations, and

frequently with a chance of success of 10-30% (Brez-Rouzaut). She quantified this

comment on the rate of return: “… oil and gas projects, where you're looking at

maybe more than 20% returns. Now you're looking at sub 10 [%] for renewables

type projects”.

The fact that the participants have an understanding of the financial aspects of

the two energy sources qualifies them to discuss the application of the cash flow

split through Royalty, Cost Recovery, Profit Sharing, Uplift, and Tax in a PSA-like

instrument.

Timing and Pace of the Transition

None of the participants were in any doubt that the transition is here, though

opinions of the timing varied from “on a fast pace” and “here to stay” to “guilty of

being a little too optimistic about how fast we’re gonna get there”, “somewhat slow in

coming” and “a good fifty-plus years to go”. It was noted that the pace varies by

region – Europe being quicker to transition than the US.

One participant reminded us that this is not the first energy transition humanity

has been through, citing the transition from coal to oil - and this researcher notes there

have also been transitions from wood to coal even earlier, and from oil to gas more

186

recently. As outlined in the “Commitment to the Environment” section of this

chapter, this transition will take the form of a phasing down of hydrocarbons and

ramping up of renewables, potentially resulting in the elimination of fossil fuels, in

several decades.

Several factors were raised in the interviews that are slowing down the transition.

These include aging transmission grids or the inability of a grid to take additional

power, technology lagging behind the desire for clean power, and the supply of rare

earth metals.

A participant talked of the common misunderstanding among the general public,

perpetuated by their leaders, that “we can flip a switch … leave all that nasty oil in the

ground… and it’s rainbows and unicorns”. Without a clear common understanding of

the issues and goals, a successful transition will be delayed. One confounding concern

that is frequently raised, and was voiced by one participant, is that renewable energy is

not truly zero emissions once building, transporting, operating, and maintaining the

facilities is factored in. Using cost and risk as an argument against transition is

considered foolish by Toh, writing for the Columbia Climate School (2021) as the

costs of unabated climate change will be far higher.

Talking Points

The first unresolved talking point concerned intermittency which is a big

concern for renewables and is discussed briefly in Chapter 2 above. The second

relates to the environmental effects of renewable energies themselves; in this

instance, the harm that offshore wind turbines might do to the bird population.

There will no doubt be negative environmental effects from renewable energy, but

187

provided those effects are less than the harm caused by hydrocarbons, there is a net

step in the right direction.

Applicability of PSAs to OWE

The discussions of whether PSA-like terms could be applied to offshore wind

yielded five positive themes and four negative ones, found in Table 27.

75% of the comments made were positive and 25% were negative.

Tone Theme Number of references

Positive Local development 23

Resource – expertise trade 15

Cost Recovery 10

Cash flow and profit-sharing 6

Drive transition 4

Negative Effect of low return on profit share 8

Ability to deliver power supply 6

Poor fit to PSA original premise 4

Legal and administrative difficulties 1

Table 27: Advantages and Disadvantages of applying PSA-like terms to OWE

Positive: Local Development

The local area benefits of using PSA-like terms for offshore wind fall into four

categories: education, training, local content, and infrastructure. Education includes

universities, which will provide the next generation of qualified staff to work on wind

farms. One participant makes the point that offshore wind will, by definition, be new

to any country without an OWF today and that providing education through the PSA

will be a benefit to the nation.

188

In terms of training, three participants saw an advantage in local workers

receiving the training they need through the PSA mechanism. The point was again

made about the OWF work being a new skill and tuition being required. Training

requirements in the PSA were seen by participants as “an attractive thing for the

Government”, “pretty popular”, and it was said that “the Government would definitely

like that aspect of it”.

Local content remarks included the potential participation of the NEC in the

project, which offers a route to learning, and a way to make a profit from an OWF.

One participant spoke of Petrobras, the Brazilian NOC, who is entering the OWE

business. This comment was verified in Memija’s (2023a) article announcing

Petrobras’s intended participation in 23 GW of OWFs. Other observations around

local content included the requirement to use local workers and facilities, thereby

developing the economy. The local workers will have been educated and trained due

to aforementioned PSA clauses.

It is probable that the local infrastructure will not yet be in place before the first

OWF is built - and that is the fourth element of local development. Three participants

suggested that the PSA could facilitate the building and maintenance of a grid that is

fit-for-purpose for the new supply from the OWF. One of those three made the point

that since the grid is built under a PSA, it belongs to the Government, which is

rightful since the country will need it for years to come, potentially long after the

Contractor has left. A second area of infrastructure arising in the interviews was port

amenities, needed for building and maintenance of offshore facilities. Again, the

comment was made that a country could be starting from scratch building

189

infrastructure for OWFs,

190

and a government would no doubt be happy to see those activities undertaken as part

of the PSA.

Overall, the remarks made on the benefits to the local economy and community

were broadly akin to those made when considering the advantages of PSAs to O&G.

Positive: Resource – Expertise Trade

One participant sees “the fit [to the PSA model] to be very good” in that “my

country has steady wind… my resource” and “you have the knowledge of how to build,

install and transmit the electricity”. Another cautioned that the wind resource would

have to be attractive enough to justify the complexity of the PSA – but as previously

discussed in this paper, there are plenty of excellent resources globally. A third

participant focussed on the advantages to the host country with no expertise. The “IOC

would have the operational knowledge… the host country would… sit back and reap

the rewards [from their resource]”.

Participants reminded us that expertise in OWFs is “new and upcoming” and

“knowledge …is …in its early stages” and it could benefit many countries if the

knowledge was transferred through a PSA-like mechanism.

NOCs, or NECs, also came up in the context of the trade, as a way to transfer the

expertise – “an important ingredient in … offshore wind”. Four participants opined

that some countries would like the opportunity to participate, for example, those with

immature NOCs, countries in Latin America, or Canada or Norway. Others such as

the USA were seen as less likely to be interested.

191

In summary, the participants saw the possibilities for PSA-like terms to provide

the same resource-expertise trade that PSAs provide for O&G.

Positive: Cost Recovery

Five participants identified the ability of the Contractor to recover costs through a

PSA-like mechanism as a positive, one calling it a “huge advantage”. To quote

another participant: “It's almost as expensive as developing an offshore oil [and] gas

project… and if you're allowed to, at the very minimum, recover your costs, I think

that would be helpful”. Risk reduction for the Contractor was mentioned several times

and two participants talked of the advantages of having a contract mechanism to

recover costs rather than just waiting for post-tax profit to outweigh investment made.

An interesting point that arose twice was that if the contract mimics a PSA

exactly, the Contractor would be awarded power in the way that it is awarded oil or

gas from a PSA in return for Cost Recovery and Profit Share oil/gas. In-depth

discussion did not ensue, and this is a point perhaps worthy of future research. Is there

any merit in the Contractor taking ownership of volumes of power which will be sold

into the local grid (export is unlikely), or should the transactions skip the resource

rights stage and be based solely on cash?

Much as PSA Cost Recovery is an advantage over other regimes in O&G, so

would it be for OWE, according to participants. Cost Recovery may perhaps be more

important for OWE, being a lower-margin business.

192

Positive: Cash Flow and Profit- Sharing

Participants saw the profit-sharing mechanism built into a PSA as a benefit for

OWE in two ways. The first is that the Contractor will have some assurance of a

return on an economic project through their profit share, taken after the costs have

been recovered. A second is that profit-sharing terms are progressive in nature and can

encourage marginal projects in a way that regressive terms such as royalties do not.

There was much discussion of profit-sharing in the context of O&G, and less

where OWE was concerned, perhaps because offshore wind is a low-margin business.

However, as renewable energy projects become more profitable, profit-sharing will

have increased utility.

Positive: Driving the Transition

One participant made the interesting point that the West is hypocritical in driving

the O&G industry throughout history, but now suggesting developing countries should

not enjoy the same level of economic development as we now know the dangers of

fossil fuels. He points out that through the PSA those countries can benefit directly

from the Western energy companies in terms of local development (as discussed

above), and this may hasten their transition.

Another thought-provoking issue raised is that developing countries using PSAs

for O&G will have familiarity and a comfort level with the contract and how it

functions. Using a similar contract may hasten their move towards OWE versus

another renewable energy with a steep learning curve attached.

193

The participants talked of mitigation of climate change as a key goal of the

energy transition and recognised that using PSA-like terms could assist.

Negative: Effect of Low Returns on Profit- Sharing

Five participants raised the issue of OWFs providing lower returns than the

“excessive returns” of a successful oil or gas field, and there being little revenue after

Cost Recovery to flow to Profit Share. One was concerned that the revenue may not

even provide for Cost Recovery, a matter that should be considered very carefully in

any investment decision, as it is for O&G projects. Another confirmed that the profit

share mechanism could work but would have to be adapted, as it is in every PSA, for

prevailing conditions. This researcher also believes that as the cost of OWE continues

to drop, a contract signed now that appears to have thin profit margins may indeed be

generating more cash than expected in two decades. Under these circumstances in the

hydrocarbon world, under a CA, the Contractor would take the upside. Under a PSA,

the host nation would enjoy the fruit of its resources. The same could be true for a

profitable future offshore wind enterprise under PSA-like terms.

A second observation was that the cash flow from the project may not sustain a

high degree of spending on non-financial benefits such as training, education, and

local content. This is another area that must be carefully considered in setting up and

negotiating a contract.

This concern for the application of profit-sharing ties to the participants’ concerns

about the financial outlook and current low returns of renewable energy projects.

194

Negative: Ability to Deliver Power Supply

Two concerns raised about the fit of offshore wind to PSA-like terms were the

ability to supply power per contract commitments, and the fact that power is not

currently exportable. The first concern will be greater when the time unit of

measurement of supply is small. If the Contractor is committed to supplying a certain

volume of power per second, or minute, and the wind is not blowing, the risk of

default is high. If the delivery is averaged over hours, days or a month, the risk

decreases. This would have to be carefully considered in the contract, in the absence

of an HRES.

The export concern is another area where the contract would differ from O&G.

Typically Contractors take their oil resource entitlement (barrels of oil equivalent to

Cost Recovery and Profit Share amounts), load it on tankers, and export it for sale.

Gas entitlement (cubic feet of gas equivalent to Cost Recovery and Profit Share

amounts) is often treated differently. It can be converted to liquified natural gas and

exported by a specialised tanker, or it may be sold locally into a pipeline network.

This latter solution informs the suggestion for power – sell locally into the grid.

Negative: Poor Fit to PSA Premise

The observations on this topic were twofold. One participant spoke of the risk

involved in offshore hydrocarbon exploration (and mentioned earlier in this paper) and

that the PSA is a good fit when large sums of risk capital are at stake. This participant

felt that there is less risk in building an OWF as the wind potential can be measured

prior to building. This is not to say a PSA cannot be used, but one of the motivations

for using it is absent.

195

The other question, raised three times, is whether the expertise to build an OWF

is as great as that needed to find hydrocarbons offshore. Participants questioned

whether the resource–expertise trade was as significant as in O&G. Neither the

researcher nor the participants are experts in OWFs and as such, may have a tendency

to underestimate the complexity of an area with which they are not familiar – the

Kruger-Dunning effect (Kruger & Dunning, 1999). It remains unknown from this

research whether an energy company is required to build an OWF or whether a

developing country government could achieve its installation. Given the big-name

energy companies that are the main players in the OWE industry, this researcher’s

suspicion is that those companies would have plenty of expertise to offer to merit the

use of a PSA.

Negative: Legal and Administrative Difficulties

This issue was only mentioned once, and mirrored comments made when

discussing O&G. All the identified negatives including complexity, lack of

transparency, and difficulty to negotiate and administrate, could apply equally to

OWE.

Opinions by Company Type

This researcher considers that analysing business opinions such as these by

demographic - gender, race, age, or income level – is unlikely to provide useful

insights. More curious is whether there are differences by company type:

consultant, IOC, or NOC. The total number of opinions given by the seven

participants on the advantages and disadvantages of PSAs in O&G and the positives

and negatives of PSAs for OWE was 228. These were split 48% by industry

Opinions of PSAs in O&G and OWE by Company

Type

Total Number of Comments

Total Positives PSA in OWE

Total Negatives PSA in OWE

Total Disadvantages PSA in O&G

Total Advantages PSA in O&G

0%10%20%30%40%50%60%70%80%90%100%

National Oil CompanyInternational Oil CompanyIndustry Consultant

196

consultants (of which there were three), 29% by IOC employees (of which there

were two), and 22% by NOCs (of which there were two). Given the small sample

size, the data were normalised by the number of participants. The consultants were

still the most talkative group, followed by IOC employees, then NOC workers, as

can be seen in Figure 24. It is true that the two NOC employees have English as

their second language, but the Norwegian, who has never lived in the USA, made

more comments than the Malaysian-American, who has lived in English-speaking

countries since adulthood. Language may affect the results, but this researcher

thinks it is more likely due to the fact that consultants are typically expected to

spend more time expressing their opinions than internal O&G company economists

(as are both NOC employees).

27% 35% 38%

31% 36% 33%

20% 27% 53%

24% 24% 53%

26% 39% 35%

Figure 24: Participant opinions on PSAs in O&G and OWE by Company Type. Data normalized by

number of participants in each group

197

When talking about the positives that PSAs could bring to OWE, the

participants were split fairly evenly across the company type at 31% NOC to 36%

IOC to 33% consultants. Similarly, when opining on the advantages of PSAs in

O&G, the split was fairly even at 29% NOC, 39% IOC, and 35% consultants. NOC

employees saw advantages of PSAs in O&G less than IOCs did, which is an

interesting result, as NOC employees may be more likely to view PSAs from the

Government’s perspective. Governments tend to like PSAs due to the option to

participate whereas energy companies can be put off by the potential interference in

daily operations (Kiyayi). It could be because the NOC participants in this research

work for a company that behaves like an IOC and does not enter PSAs on the side

of the Government but on the side of the IOC.

When it comes to the negatives, consultants came to the fore. Regarding both

the disadvantages of PSAs in O&G and the negatives of PSAs in OWE, consultants

voiced over half the opinions (53% for both). This researcher can only speculate at

why this is, but perhaps it has to do with consultants being career problem-solvers,

and identifying negatives may come naturally and foremost to them. An implication

is that if this concept – PSA-like terms in OWE – is to be introduced to potential

users, it should be demonstrated to the IOCs and NOCs (and governments) directly.

It should not be introduced through consultants who might portray it negatively.

Had government or energy ministry employees been included in the study,

these results might have been different. It is possible they would have seen more

positives of PSAs. This is analysis that could be undertaken in future if the

participant sample is expanded.

198

CHAPTER 5

DISCUSSION

This study is timely as the effects of climate change and the importance of efforts

to mitigate it are widely and frequently publicised. The research takes an O&G best

practice – the PSA - and applies it to OWE. This provides an additional option for the

contract governing the leasing of seabed for OWFs over and above the contracts used

today in the few countries with offshore wind. It also further expands the list of the

O&G practices being harnessed by offshore wind as discussed in “Learning from the

Oil and Gas Industry” in Chapter 2.

The research is aimed primarily at developing countries which have OWE

resources but no OWFs. Offshore wind was not an accessible solution for such

countries when it was a high-cost energy source and needed government subsidisation.

The path may not be smooth or linear but OWE costs are lower, competitive, and

continuing to fall, and the days of subsidisation are passing or over (Jansen et al,

2022). Developing countries are starting to take an interest in OWE, including

Vietnam, India, Brazil (discussed in Chapter 2), Sri Lanka (Boopathi et al, 2023,

Economynext, 2023), Bangladesh (Garg, 2023), and Colombia (Anchustegui, 2023;

Emmanuel). Of these countries, all but Colombia use PSAs and enjoy the benefits

identified in the qualitative research.

This convergent mixed method research was conducted in two parallel phases and

found that PSA-like terms could successfully be applied to OWE. Success is defined

quantitatively by project economic viability and qualitatively by the provision of

benefits to the parties to the agreement.

199

Economic Viability and Optimal Contract Terms

It was concluded that under the contract terms, discount rates, generation, price,

and cost conditions used in this research, PSA-like terms can be used as a leasing

instrument for acreage on which to site an OWF. Pursuant to T-tests, the null

hypotheses associated with H1-1 and H1-2 were rejected and both Government and

Contractor mean NPVs are statistically greater than zero (RQ1). Since both

contracting parties will realise a positive NPV they should rationally be willing to

move forward with the project, in the absence of a better project competing for

financial or human capital. This research is the first to apply the terms to OWE and is

also unusual in considering both the Contractor and Government outcomes in the

meta-modelling of a PSA.

The sensitivity of the NPVs to the fiscal terms was assessed broadly through

multiple regression and the models yielded good explanatory power with R2 of 84%

(Contractor) and 90% (Government) (RQ2 and 3). The null hypotheses associated

with hypotheses 2-1-1 through 2-2-5 and 3-1-1 through 3-2-2 were rejected following

the multiple regression analysis. The contract terms associated with those hypotheses:

Royalty, CRC, CPS, Uplift Rate, Discount Rate, Tax Rate, Average Price, and

Average Generation all have a significant effect on NPV.

Using scenario analysis, the behaviour of the five fiscal contract term IVs was

studied, and various idiosyncrasies uncovered. Further, the sensitivity of each term

within a small range was tested using the same technique, which would be useful

during negotiation to uncover the optimal set of terms (RQ4). A user of the model can

elect to constrain variables to small ranges or allow the model to draw from the full

200

distribution. For example, the user could study a small range of fiscal terms in

combination with the full distribution of Price and Generation data, which has utility

in real-world negotiations. This research found that the total NPV is somewhat

insensitive to changes in Price, Cost, and Generation, but a successful project outcome

(positive NPVs for both contract parties) is more sensitive to changes in those IVs

(RQ5). This is important information when selecting a site for an OWF and designing

the turbines and their placement. Further, knowing the sensitivity of project

economics to Price is helpful in deciding whether to seek a PPA which locks in the

power sales price. This long-term price guarantee is not something available for oil

production.

Multiple regression and stochastic scenario analysis are advanced techniques and

could complement the single-point NPVs and deterministic scenario analysis often

seen in the O&G industry. It would be interesting in the future to perform the same

analyses for an oil or gas field, and the conclusions would have utility for Contractors

or Governments negotiating O&G leases. The cash flow model can be adapted to

such that the technical inputs can be specific to a certain oil or gas field and the terms

represent a specific PSA.

Other future directions for this quantitative research include studying other more

minor fiscal terms found in PSAs. These will have less impact on NPV but

nonetheless exist in O&G and may have utility in OWE. Likewise, the dependent

variable list could be increased to include more economic indicators, such as IRR and

Contractor and Government Take, as studied by KP, and profitability indicators,

popularly calculated in the industry.

201

Advantages and Disadvantages of PSAs in O&G and Applicability to OWE

Many of the themes arising from the interviews were common to both O&G and

OWE discussions. Figure 25 shows the three themes that are common among the

advantages, and the three O&G themes and one offshore wind theme that are not

common.

202

Figure 25: The common and unique themes when comparing PSA advantages across O&G and OWE

203

The non-financial benefits of PSAs; the education, training, jobs, infrastructure,

and participation of the NOC in the project were seen as important in O&G and of

great potential benefit for OWE if the project economics can sustain the expense.

Focussing briefly on the last, NOCs, Strambo et al (2023) studied four Latin

American NOCs for their strategic direction with regards to climate change and found

challenges. The authors point to the existential threat to NOCs who focus solely on

hydrocarbons as the world transitions to renewables. They also point to the immense

opportunity NOCs have to drive change, as they own 58-66% of the world’s

hydrocarbons, but identify two challenges that they face. Those are limited capacity

for change and institutional barriers. Should OWE come with all the above-mentioned

advantages, perhaps it would be easier for NOCs to change strategy and protect their

very existence.

Cost recovery, viewed as a tremendous and unique feature of the PSA, could

have even greater importance in expensive OWE projects with low returns. Profit-

sharing is also perceived as an excellent contract term in O&G due to its progressive

nature and ability to reward the resource owner in highly profitable projects. This too

has application to OWE especially as projects are likely to become more profitable in

the future. Governments who establish a limited share of an OWF profit today,

thinking that profits will be low, may regret that decision in the future if profits rise

and all flow to the Contractor.

The trade of expertise for resources and the knowledge transfer that comes with

the PSA structure and the relationships it governs works very well in O&G and is

similarly applicable to OWE.

204

The three O&G themes not raised in the OWE discussion could still have

applicability – if the contract was structured comparably. The host would retain

ownership of the resources, would have low-cost, low-risk participation, and the PSA

would be an enduring contract between the parties. Perhaps given more interviews

these would also have been identified as OWE themes. The last unique theme, driving

the transition, is clearly unique to OWE.

Similarly, there was commonality among the disadvantages of PSAs across O&G

and OWE, shown in Figure 26. Two themes were in common, and each of O&G and

OWE had two unique themes.

205

Figure 26: The common and unique themes when comparing PSA disadvantages across O&G and

OWE

The difficulty associated with PSAs being complex to understand and administer

in O&G would clearly be relevant to OWE and was identified as such. The limitation

of Contractor upside in O&G is not the same issue as the ineffectiveness of profit-

sharing in low-return environments but affects the same fiscal term. In O&G the issue

was felt to be the capping of upside in highly profitable projects and the awarding of

the majority of the profits to the Government. However, low-profit projects can occur

in either O&G or OWE, then the point made about OWE - that there is little profit to

206

be shared – is equally relevant. However, the opinions of the participants about the

flexibility of profit-sharing and its progressive nature are applicable regardless of the

energy source. In the future, as costs fall further, OWFs could be characterised by a

wide range of profitability, and then PSA-like profit-sharing would offer the same

flexibility that it does in O&G.

O&G’s first unique theme - abuse or instability of contract - could equally be a

disadvantage to OWE but was not mentioned in the interviews. The second –

Contractor entitlement in kind – is a moot point. The Contractor is unlikely to want to

export power, as previously discussed, but the contract could allow for this should it

wish to.

OWE’s first unique disadvantage – the poor fit of a few PSA terms to some

facets of offshore wind – is applicable only in that arena. The second – the ability to

deliver power supply – is certainly an issue in a one-source power generation system

but could be mitigated, as argued elsewhere in this paper, by an HRES or by carefully

selecting the periodicity of measurement of supply.

In summary, the participants found many positives about the use of PSAs for

O&G, in fact, positive references outweighed negatives 56 to 18. Many of these

positives can also make a PSA-like contract appropriate for OWE. Of the negatives

about PSAs in O&G, most would also be disadvantages for PSA-like terms in OWE.

The PSA for offshore wind would have to be carefully crafted to take into account the

differences between the two energy sources, such as today’s low-profit margins of

wind and the fact all the power generation is likely to be supplied to the local grid.

207

Future research should look in more detail at specific PSA terms and how they

should be adapted. Open questions following the interviews may also produce

valuable future research. These included periodicity of supply commitment and

intermittency issues, Contractor entitlement to power or cash, rights to sell into the

local grid, and whether a discounted DMO for the host country is appropriate or

feasible in a low-margin business.

Further, expanding the sample and conducting more interviews, and

incorporating Government or energy Ministry employees would be worthwhile - given

the significant time needed to obtain IRB approval for non-USA participants.

Conclusion

PSA-like contract terms could be applied to a profit-making OWF and provide

another style of lease contract between OWF developers and the resource owner.

The contract fiscal terms can fit an OWF in multiple ways. The lower Royalty found

in a PSA versus a CA regime can encourage marginal projects better than a high

Royalty. Early OWFs without subsidisation may suffer marginal returns. The Cost

Recovery mechanism can encourage Contractors to take the risk of leasing seabed

and building OWFs. The profit-sharing mechanism, as in O&G, can allow the

Contractor a reasonable rate of return but award the bulk of the excess cash flow in a

profitable project to the resource owner, the Government. Even if the Contractor is

not in a taxable position in the country and will pay no tax, the Government will

benefit from profit-sharing. As the PSA is a contract, and not a regime covering all

offshore wind activities in a given country, the details of the Profit Split can be

tailored to the individual project or area. The Uplift mechanism can be employed

much as it is in

208

O&G for new technologies and new areas. The new technology here is offshore wind

and the new geography is developing countries.

The non-financial terms of the contract will benefit the local economy which will

see regional industry and infrastructure developed and jobs created. The Contractor

will benefit from the improvements they make to the infrastructure that they need for

the project. Finally, the Government may participate in the project, setting up an NEC,

and learning to maintain the project or build new ones after the Contractor has moved

on. With these advantages available, a developing country using PSAs for O&G may

find the switch to OWE attractive and may even be willing to forgo some hydrocarbon

development for advances in a cleaner energy source. Since no developing countries

except Vietnam have any offshore wind capacity and few others are in the planning

phase, the timing is right to introduce a new contract type.

In order to use a PSA-like contract, a government would need to design the

detailed terms and offer acreage to the industry under this new contract. Bidding for,

or direct negotiation of, the terms could ensue, and the Government would partner

with its preferred candidate and sign the contract. The Government should be mindful

of the negatives associated with the contract, such as complexity, and design it

carefully. The cash flow model used in this research could assist both parties with

negotiation and planning.

Namibia would be an interesting candidate for this new contract type. As stated

before, the country imports almost all the power it requires (Choti & Xydis). It does

not have O&G production off its shores but had its first oil discovery in 2023 (S&P

Global, 2023) and expects first oil production in 2030 (Roelf, 2023). A transition to

energy independence not through oil but through offshore wind would be less

damaging to the climate, and Namibia’s wind potential is among the best (Figure 1 &

Figure 2).

Future research on this topic could expand the PSA-like contract to other sources

of renewable energy. Other research could analyse the situation and applicability in

specific developing countries, knowing the wind potential, likely costs, market, and

grid availability, infrastructure situation, and the terms currently in use in their PSAs

in O&G.

Finally, since PSA stands for Production Sharing Agreement, and this new

contract will neither have production (refers to O&G) nor the in-kind entitlement

clause which awards the Contractor barrels of oil or cubic feet of gas, a new name is

proposed by this researcher: “Offshore Wind Acreage Usage Agreement”.

Exploring the Applicability of Production Sharing Agreements to Offshore Wind Projects

Students also viewed