Urban Energy Master Planning Methodology for District Energy Systems

This thesis examines how we can best plan a district energy system in an

urban area to meet energy and emission reduction goals. This is a timely question as

many urban areas have fast-approaching climate action goals, and current planning

methods fail to account for the specific challenges urban communities face.

1.1 Urban Energy

Urban areas consume seventy-eight percent of the world’s energy and produce

more than sixty percent of the greenhouse gas emissions [1]. These numbers are

expected to increase as the urban population is predicted to double by 2050 [2].

Urban areas also have ambitious energy and carbon reduction goals. Below are some

examples of major urban areas and their climate action plans:

•Austin TX, Austin Community Climate Plan [3] o Net-zero greenhouse gas

emissions by 2050 o 55% renewable energy by 2025

•Denver CO, Denver 80 x 50 Climate Action Plan [3] o 80% reduction in

greenhouse gas emissions by 2050 from a 2005 baseline o 100% renewable

electricity usage in the community by 2030 o 50% reduction in commercial

building energy by 2050 o 100% renewable electricity usage in municipal

buildings by 2025

•Los Angeles CA, L.A.'s Green New Deal [3] o 100% renewable energy by

2045 o carbon neutral by 2050

Currently, renewable energy initiatives are one of the best ways to achieve climate

action plans quickly; however, this approach can present many challenges for urban

areas. Some of the challenges urban areas face include:

•More significant seasonal energy load swings due to heat island effects and

climate change

•Environmental and noise regulations

•Limited space

•Utility restrictions on how energy can be generated, stored, and sold

•Increased cost burden, as city buildings typically use more energy and require

larger equipment to support it

These challenges impact the type and size of energy generation and storage

technology that can be used in an urban community and how they can be operated.

The impact of these challenges will determine how successful an urban community

can be at meeting its climate action plans. District energy systems are one-way urban

communities can successfully approach these challenges by allowing urban areas to

share resources and make the most of energy technologies while decreasing their cost

and vulnerability.

1.2 District Energy Systems

District energy systems (DESs) are an efficient and reliable way to provide

electricity, heating, and cooling to multiple buildings in a central location. They are

implemented through the creation of a distributed energy network, as seen in Figure

1. Electricity for a DES can be produced in various ways, including combined heat

and power

(CHP), renewable energy technologies, and gas generators. Likewise, heating (steam

or hot water) and cooling (chilled water) for a DES can also be produced in various

ways, including CHP, boilers, geothermal, chillers, and cooling towers. Both battery

and thermal storage (hot and cold) can be integrated into a DES to help shift loads to

off-peak times, balance load supply and demand, and support a system in an

emergency [4]. Most DESs have a mixture of multiple energy sources, which can

improve the system’s reliability and resiliency.

Figure 1. Example of a district energy system (DES) illustration showcasing the

ability of the central plant to support multiple buildings through a distribution

network. The illustration also shows how a DES can utilize a variety of energy

generation and storage technologies.

DESs can support various building user groups such as cities, college

campuses, healthcare facilities, and military bases. Currently, there are more than 700

DESs in the U.S. [5]. DESs offer the possibility for each of these groups to become a

microgrid, and, with the support of renewable technologies and energy storage,

become self-sufficient. This allows critical infrastructure within these groups to stay

operational during an emergency. A district energy approach allows building use

groups to have reduced operating costs, lowered greenhouse gas emissions, and

decreased peak electricity demand compared to systems that get all their electricity

from the grid [6]. These benefits save money, improve energy efficiency, and

decrease system vulnerabilities. Examples of current DESs can be seen in Table 1.

However, DESs can be challenging to plan and implement, especially when

integrated into an existing developed urban area. Plans must account for space

restrictions, funding opportunities, environmental and policy regulations, load

fluctuations, equipment options and performance, and the ability to withstand

emergencies. Each DES is unique and must be customized to each building group’s

needs, and when designed and operated appropriately the DES will optimize the

system’s impact. One way to overcome the design and implementation challenges

and ensure system optimization is to use a building energy modeling tool.

Table 1. Examples of district energy systems.

Name,

Location

Energy

Loads

Supported

Number

Buildings

Supported

Type of

Buildings

Supported

Footprint Plant

Equipment

Fuel(s)

Used

Capitol

Power

Plant,

Washington

D.C. [5,7,8] 1

Electricity,

steam, and

chilled

water

buildings

for steam

and 19

buildings

for chilled

water,

electricity

sold to the

grid

Gov office

buildings

(primary),

conservatory

(botanical

gardens),

transportation

(train station),

and a data

center

Supports

18.4 million

sqft of

facilities,

570 acres of

grounds,

and serves

30,000

occupants

daily

7 boilers

(620,000

lbs/hr), 9

chillers (39,600

tons), and 1

CHP (7.5 MW)

Natural

gas, fuel

oil, and

coal

Princeton

University,

Princeton

NJ [5,9,10] 2

Electricity,

steam, and

chilled

water

150

buildings

for steam,

chilled

water, and

electricity

Academic,

research,

administrative,

residential, and

athletic

Supports

9.5 million

sqft of

facilities,

and serves

8,000

students and

employees

1 HRSG

(180,000 pph),

2 boilers

(300,000 pph),

3 steam driven

chillers (10,100

tons), 5 chillers

(10,700), 1 CHP

(14.6 MW), and

16,500 solar

panels (5.3

MW)

Natural

gas, fuel

oil, and

biodiesel

1 The author would like to thank and acknowledge Chris Potter, Director of Utilities and Power Plant Operations at

the Architect of the Capitol for providing information on their DES.

2 The author would like to thank and acknowledge Ted Borer, Energy Plant Manager at Princeton for providing

information on their DES.

New York

University,

New York

City NY

[5,11,12]

Electricity,

steam, and

chilled

water

buildings

for heat

and chilled

water, and

buildings

for

electricity

Academic,

research,

residential,

office, campus

operations,

athletic,

commercial,

health services

Supports 5

million sqft

of facilities,

and 67,000

students and

employees

Boilers and

HRSG (113,240

pph), chillers

(10,000 tons),

1 CHP (5.5 MW

GT and 2.4 MW

ST)

Natural

gas, and

fuel oil

Carl J.

Eckhardt

Combined

Heating and

Power

Complex at

UT Austin,

Austin TX

[13,14,15] 3

Electricity,

steam, and

chilled

water

160

buildings

(whole

campus)

for heating,

cooling,

and

electricity

Academic,

research,

residential,

office, campus

operations,

athletic,

commercial,

health

services, and

chilled water

to Dell Seton

Medical

Center (a

hospital)

Supports 22

million sqft

of facilities,

and 55,000

students and

employees

Boilers and

HRSG (1.2

million pph),

chillers (60,000

tons), 1 CHP

(135 MW),

solar panels

(546 kW), and 2

chilled water

thermal energy

storage tanks

(80,000

tonhours)

Natural

gas, and

fuel oil

Harvard

University

Main

Campus

and District

Energy

Facility

(DEF),

Cambridge

and Allston

[16,17,18,19,20] 4

Electricity,

steam, hot

water, and

chilled

water

160

buildings

across both

campuses

Academic,

research,

residential,

office, campus

operations,

athletic,

commercial,

health services

Supports 21

million sqft

of facilities

and 41,000

students and

faculty

across both

campuses

Main Campus:

Two chiller

plant (20,500

tons), four

boilers and 1

HRSG, 1 CHP

(7.5 MW GT

and 5 MW ST),

and solar panels

(1 MW)

DEF Campus:

chiller plant

(15,000 tons),

1chilled water

thermal energy

Natural

gas, and

fuel oil

3 The author would like to thank and acknowledge Juan Ontiveros, Vice President for Utilities and Facilities

Management at the University of Texas at Austin for providing information on their DES.

4 The author would like to thank and acknowledge Bob Manning, Director of Engineering and Utilities at Harvard

University for providing information on their DES.

storage tank

(1.3 million

gallon), boilers

(14,000 MBH

hot water), 1

CHP (7.5 MW)

Chapter 2: Modeling the Built Environment

2.1 Building Energy Modeling

Building energy modeling (BEM) is an effective approach to representing and

understanding a building’s design and operation. BEM can be used to improve a

building’s energy efficiency, reduce heating and cooling loads, and analyze how a

building will perform under specific stresses. It can also promote and inform new

policies and regulations, which can support the economy and lower greenhouse gas

emissions. Currently, there are over 820 different BEM tools available [21]. Each tool

focuses on a different aspect of modeling buildings. Some BEMs focus on HVAC

system selection and sizing, or on building code compliance and Leadership in

Energy and Environmental Design certification, while others focus on whole building

simulation and analysis. These tools have been developed by industry, academia, and

the U.S. Department of Energy national laboratories. They all have benefits and

limitations, but each tool offers unique capabilities and plays an important role in

reducing the vast amount of energy consumed, and emissions produced by buildings.

BEM tools are currently the industry standard for the master planning of a building,

including its energy systems.

When modeling a DES, bottom-up BEM approaches (data-driven models,

physical models, or a hybrid of the two) are used [22]. Bottom-up approaches are

more adaptive and predictive of energy system changes compared to statistical and

analytical model approaches (top-down) [23]. Data-driven models use real energy

system data (i.e., energy loads, equipment schedules, plant performance, occupancy

levels) to perform statistical regressions and calculate building energy consumption

without loss [24]. While datadriven models are reasonably accurate, they are often

hard to scale to multiple buildings and do not integrate DES interconnections well

[25]. Physical models are the most common method of modeling DESs [26]. Physical

models use physical data about the building and building equipment (i.e., volumetric

flow rates, heat transfer coefficients, rated electricity, and wall thickness) to

determine the energy and mass balance of the energy system. This modeling

approach requires the user to have an in-depth understanding of the building, and it is

not always accurate as equipment and plants do not always operate at their rated

specifications. These models do not consider how the building(s) is performing or

being operated, which can cause significant errors in the modeling results. Hybrid

models are a combination of a data-driven and physical model. They aim to use real

system data to help improve the accuracy of the physical model and give more

capability to the data-driven model [24]. However, they are computationally

expensive and complicated (due to combining two different approaches), so they

have limited scalability.

2.2 District Energy System Modeling

A BEM must have certain key attributes for it to accurately model a DES and

for both the BEM and non-BEM communities to easily use and understand the

model. These

attributes are:

•Co-simulation of buildings/energy systems

•Component/Control flexibility

•Individual building specifications

•Multiple building loads

•CHP/Power generation equipment

•Renewable generation technologies

•Energy storage technologies

•Microgrid/Emergency capabilities

•Lifecycle/Cost analysis

•Graphical user interface

The tool should model each building within the DES with an appropriate amount of

detail, including each building’s envelope, loads, equipment, HVAC loops, and

thermal zones. By including individual building specifications, the model becomes

more accurate, and the user gains greater insight into the DES's performance.

Additionally, the tool should allow the user to define specific system controls and

measures. This allows for further system customization necessary to represent the

variations within DESs. The software needs to model different types of equipment,

including those typically used for HVAC and power generation (e.g., chillers, cooling

towers, steam generators, boilers, and gas turbines). Each of these pieces of

equipment plays a critical role in supplying energy to a DES. Thus, each piece is

important to model and understand. The software should also allow users to adjust

equipment specifications and schedules, such as the steam production capacity of the

CHP, the maximum condenser water flowrate, and what time of the day the chillers

are running. This is important because every DES is unique and requires different

equipment specifications and schedules. The available equipment should also expand

to renewable generation and energy storage, as both are key components of a next-

generation DESs.

Renewable generation and energy storage diversify a DES, making it more

reliable and resilient, allowing it to function during an emergency. However,

renewable generation and energy storage are relatively new technologies and have

not had the time to fully develop and integrate into all BEM tools. Since DESs

support multiple buildings, the BEM should be able to model each of these buildings

and simulate all the buildings and energy systems at once, otherwise referred to as

co-simulation [26]. Co-simulation is important because the goal of using a BEM for a

DES is to capture the complex and dynamic interdependencies and interconnections

between each building and the environment [27]. Finally, the tool should be user-

friendly with an easy-to-use graphical user interface, and the simulation should

produce easy-to-read graphs and data tables showing the model

results.

Currently, no BEM tool meets all the attributes described above. This is

mainly because BEM tools are not widely used in the DES industry and thus are still

immature [28]. Table 2 shows five existing BEM tools currently used by the energy

industry to model and simulate DESs. Each BEM software is evaluated on its ability

to meet the ten key attributes (listed previously). These attributes highlight each

software tool’s strengths and weaknesses and show the limitations of using this

software in modeling and simulating DESs. With improved BEM software,

companies and government organizations can better understand and demonstrate a

DES’s impact on meeting an urban community’s climate

action goals.

Table 2. Summary of tools used for building energy modeling and their ability to

capture key district energy system attributes.

Attributes District Energy Modeling Tools

URBANopt

™ [29, 30] 5

OpenBlue Plant

Simulator™

[31, 32] 6

MATLAB® -

Simulink® and

Simscape™

libraries [33]

OpenModelica®

- Modelica®

Buildings

Library [34, 35]

HOMER

Pro® [36]

Developer NREL Johnson Controls

Inc

MathWorks® LBNL Homer

Energy

5 The author would like to thank and acknowledge Ben Polly, Senior Research Engineer at the National Renewable

Energy Laboratory for providing information on URBANopt.

6 The author would like to thank and acknowledge Derek Supple, Product General Manager and Director of Product

Management at Johnson Controls Inc for providing information and access to the OpenBlue Plant Simulator.

Open Source/Free Yes Yes, with certain

options only

available through

purchase

No Yes No

Can model renewables Not yet Yes Yes Yes Yes

Can model battery and

thermal storage

Not yet Just thermal Yes Yes Just battery

Can model multiple

building loads

Yes No Yes Yes No

Can model building

envelopes and account

for individual building

equipment and

specifications

Yes No No Yes No

Can model a microgrid Not yet Yes Yes Yes – with help

from other

libraries

Yes

Can model a CHP and

power generation

equipment

Not yet Yes Yes Yes Yes, except

for CHP

Can perform lifecycle

and cost analysis

Not yet Yes No No Yes

Allows the user to create

system

components and

controls not defined by

the software

Yes No Yes Yes Yes –

through

MATLAB

Can co-simulate

multiple buildings and

energy systems

Not yet No No No No

Graphical user interface No Yes Yes Yes Yes

2.2.1 URBANopt™

URBANopt™ is a promising BEM tool being developed out of the National

Renewable Energy Laboratory (NREL). It is based on NREL’s current BEM software

development kit, OpenStudio®, and NREL’s whole building simulation engine,

EnergyPlus®. OpenStudio® is an open-source whole-building energy modeling tool

[29].

When coupled with EnergyPlus®, OpenStudio® can model and simulate a building’s

envelope, loads, HVAC system, equipment schedules, and thermal zones.

URBANopt™ will utilize the Modelica building library and building language, which

was developed by the Lawrence Berkeley National Laboratory to model HVAC

controls, multizone airflow, computational fluid dynamics related to building heat

transfer and HVAC systems, and load prediction. URBANopt™ will also make use of

the Open Distribution System Simulator (OpenDSS), the Renewable Energy

Integration and Optimization (REopt™) techno-economic decision support platform,

and the next-generation simulation engine, Spawn-of-EnergyPlus [29]. OpenDSS is

an electrical simulation tool focused on analyzing grid modernization and distributed

resource integration techniques. REopt™ evaluates energy generation and storage

technologies based on cost, resilience, and energy performance goals. Finally,

Spawn-of-EnergyPlus uses a dynamic state-based simulation to execute building

system control sequences that can be applied to physical building

controllers.

URBANopt™ is still under development, but when finished, it will be able to

model full DESs with the same detail and analysis that OpenStudio® provides for

modeling singular buildings. URBANopt™ will allow for the design and analysis of

physically realistic thermal and electrical systems and the co-simulation and control

of multiple buildings. This tool is extremely customizable and will enable users to

adjust, specify, and control every aspect of a DES. For example, users will be able to

input air flow rates and energy loads of individual rooms within a building, hourly

water temperature and flowrates through system components, and energy losses from

network connections. The highly detailed approach improves the model’s accuracy

and allows for the modeling of any DES. It gives the user a deep understanding of the

modeled system and simulated scenarios. By modeling each sensor or meter in the

actual system, URBANopt™ models produce data that can be compared directly to the

actual system’s operational data to find system process or equipment errors; the

models further allow the user to make changes to the modeled system to predict the

projected impact on the existing system. However, this approach requires immense

expertise in both building systems and the software itself, making the modeling

process time-consuming and not user-friendly to people outside the development

team or with highly specialized expertise. There are also limited resources to help

people, especially novices, learn how to use this modeling tool and mitigate modeling

errors.

2.2.2 OpenBlue Plant Simulator™

Johnson Controls Inc. created the OpenBlue Plant Simulator™ to accompany

their system operational control software. This modeling tool supports the design and

planning of new central utility plants and plant upgrades [31]. It allows users to easily

compare multiple equipment specifications, plant configurations, and different

technology options. It uses a model predictive control strategy to simulate hourly

utility rates and usage for one year. This tool focuses on the details of the central

utility plant and the equipment it supports [32].

This modeling tool allows the user to easily create a DES with multiple

energy generation and storage sources, as seen in Figure 2. The ability to model

battery storage is still under development. This tool also lets the user customize

equipment specifications and controls. However, the tool cannot model individual

buildings. Instead, it captures the system’s total electricity, heating, and cooling loads.

While this enables quick modeling and simulation, it limits the analysis that can be

conducted and the understanding of the DES as both a whole and as a network of

individual buildings. Thus, the models cannot account for individual building

specifications, individual building equipment, and energy losses from building

connections, all of which are important parts of a DES.

Figure 2. An image of the Johnson Controls Inc. OpenBlue Plant Simulator™ interface

showing a district energy system model. Where nodes are used to connect system

components and control the flow of resources (steam, natural gas, water, electricity,

and exhaust gas) between components, and lines represent the flow of resources

between system components.

2.2.3 MATLAB® with the Simulink® and Simscape™ Libraries

Simulink® software and Simscape™ software are libraries in the MATLAB®

family of products; they were developed by MathWorks. These libraries are designed

to model and simulate system-level designs and components. They use simple blocks

that represent inputs and outputs through mathematical functions and signals that

connect blocks and transfer values between them in the form of vectors and matrices.

These simple blocks and signals allow the user to build various equipment and

systems, such as a CHP plant, as seen in Figure 3. Currently, many predefined system

components exist, such as renewable generation and storage technologies. However,

these libraries focus primarily on the controls and signals of a system and lose

precision and robustness when simulating a system’s physical properties, such as

flow rate, temperature changes, and thermal losses [33].

Figure 3. Depiction of a combined heat and power plant modeled with MATLAB®

software. The image shows the many inputs and system components required for one

component of a district energy system.

While the Simulink® software and Simscape™ software have been used to

model building energy systems, this was not their intended purpose; thus, the models

are simple and lack the functionality to incorporate system details and the physical

analysis of other modeling platforms. This tool also requires the user to have a deep

knowledge of mathematical modeling and understand how control theory is applied

to the design of electromechanical systems. Significantly, this software allows for

hardware-in-the-loop testing and verification, which bridges the gap between

modeling and implementing a district energy system.

2.2.4 OpenModelica with the Modelica Buildings Library

The Modelica Buildings® library is a physics-based modeling library

developed by the Lawrence Berkley National Laboratory. It is an open-source library

designed for building and district energy control systems. The platform allows the

user to make simple component-based models, which can then be simulated in

OpenModelica® or imported into a platform such as Simulink® for simulation and

analysis [33]. This library allows for whole-building energy modeling, including

individual building envelopes, HVAC systems, and thermal zones. The library is also

equipped to model CHP components, renewable technologies, and energy storage.

This tool is a good option for modeling new and current building systems, and it can

be used to design and test equipment specifications and control sequences [35].

The component-based approach is ideal for modeling small systems, but it

can be quite complex when modeling large district energy systems because every

component and controller must be imported and designed by the user. This approach

mixes a graphical user interface with components that can be dragged into the model,

as seen in Figure 4, and code that the user writes to set equipment specifications and

control sequences, as seen in Figure 5. This structure results in a detailed analysis but

makes the tool less user-friendly because there are many variables and points where

errors can arise.

Figure 4. Model diagram of a cooling system in OpenModelica®.

Figure 5. Code cell in OpenModelica® used to define equipment specifications and

control sequences.

2.2.5 HOMER Pro®

HOMER Pro® is a simple and extremely user-friendly modeling tool designed

for the economic optimization of microgrids. However, it is not a physics-based

modeling tool, and it focuses heavily on the electrical aspects of the energy system. It

cannot model individual buildings or building specifications and instead considers

the primary electrical and thermal loads. The defined thermal loads are limited to heat

supplied by a boiler, a heat recovery steam generator, or heat caused by an “excess of

electricity” (although the latter is not defined in more detail). However, these are not

the only heating sources in a DES. Other common heating sources include

geothermal, solar thermal, steam generators, and heat pumps. Notably, the tool cannot

model cooling loads [36]. Cooling loads are an important part of a DES as coolant is

chilled in a central location and distributed to all the buildings within the DES

network. Rather than having cooling generated by electricitypowered air conditioning

units, cooling in a DES uses chilled water, produced by chillers, cooling towers, or a

geothermal system.

This tool’s simplicity allows for rapid modeling and the simulation of

multiple scenarios easily. It can determine optimal system configurations as well as

net present, capital, and O&M costs [37]. Its graphical user interface, seen in Figure

6, is straightforward, producing graphs of the data that are easy to understand.

Figure 6. An image of the HOMER Pro® application. In this application users can

easily select system attributes and choose how the system will be analyzed.

2.3 Building Energy Modeling Software Limitations

There are currently many BEM software limitations to modeling district

energy systems [38]. One major limitation of current BEM tools is their inability to

model a network of buildings and the equipment that supports them while

maintaining the same detail currently available in singular building modeling.

Furthermore, BEM software tools cannot co-simulate buildings within the DES.

Because of these two limitations, BEM users must model and simulate individual

buildings and then sum all the resulting building loads. This summation produces

primary electrical, heating, and cooling loads. These loads can then be applied to a

different BEM software capable of modeling and simulating a distributed energy

supply network. This approach leads to errors in the simulated results because energy

losses caused by system network connections are not captured. Additionally, using

multiple BEM software introduces the chance for human error when transferring data

from one system to the next, especially if the data needs to be manipulated to fit the

new system (e.g., changing data units).

Another limitation of BEM software is the variety of equipment that can be

currently modeled. There is a time gap between when a technology is implemented

into an energy system or DES and when it can be modeled using software. For

example, grid-scale batteries were first introduced into the U.S. market in 2009 [39].

Yet, they are just now beginning to be successfully modeled in BEM software, and

some companies are still developing the ability to model batteries. Another limitation

is the variety of batteries that can be modeled. Most BEM software has a minimal

selection in the type of battery that can be modeled and simulated. This is a limitation

because it restricts how a user can define a DES and it may not allow the user to

model their system fully, especially if their system contains a piece of equipment not

currently supported by BEM software. A limited variety of equipment is also a

limitation because the user cannot model the newest energy technologies and

advancements. New technologies are often the most interesting and relevant for

modeling because they allow the user to understand the impact of these new

technologies and advancements on their system.

Additionally, modeling tools are limited by the amount of knowledge and

accuracy the user has on each building being modeled. Every BEM tool has many

inputs, which require a vast amount of data from each modeled building. This

building data can sometimes be obtained from architectural drawings, audits, or

meters; however, this data may not be available to the user due to a lack of metering

or data integrity issues, especially as the model incorporates more buildings within

the DES. When data is unknown, the user of the BEM tool is forced to assume

values, which limits the accuracy of the modeling results [40].

The next limitation of BEM software is the inability to account for occupancy

changes and human behavior factors. The unpredictability of occupants is one of the

main reasons for the error between modeled and actual energy consumption [41].

Human building interactions (HBIs), such as changes in temperature setpoints,

lighting systems, plug loads, and opening doors and windows, all greatly impact

energy consumption [42]. However, HBIs are incredibly difficult to model, and

attempts to do so through machine learning, stochastic representations [43], or

parametric schedules [44] are time-consuming and computationally expensive [45].

The final limitation is the user-friendliness of BEM software. The current

software tools showing the most progress towards fully modeling a DES are hard to

use and require a deep technical knowledge of mechanical, electrical, and control

systems. While the primary focus of development should be on the technical side,

developers should not dismiss the importance of user-friendliness. Adoption of a

modeling tool in the building energy sector will be accelerated if the tool is easy to

learn and use, has a good graphical user interface, and produces simulation results in

an understandable format. These modeling tools offer a gateway for government and

industry organizations to learn about DESs and evaluate the impact of a DES in their

communities. Difficult software limits the impact of a BEM tool on the progression

of DES implementation and advancement.

While building energy modeling tools have made great progress over the

recent years, they still have limitations to overcome before they can fully model a

district energy system. Currently, many different modeling tools are available beyond

those discussed here, and they each have attributes relevant to modeling the many

components and requirements of a district energy system. While the approach of

using multiple building energy modeling tools works, and will be used in this paper,

there is a need for a tool that simplifies this process and increases its accuracy. There

is also a lack of cohesiveness within the building energy modeling community. The

community would benefit from some standards or set approaches to what modeling

tools should be used and how their simulation results should be evaluated.

Additionally, developing educational and professional programs around using BEM

tools could expand the knowledge base within the community and help advance DES

implementation.

Chapter 3: District Energy Case Studies

The following chapter discusses two district energy systems, which will act as

case studies for the following chapters. The first case study is a synthetic DES, a

representative urban community. The second case study is of The George Washington

University’s central utility plant (CUP), a real urban community.

3.1 Synthetic District Energy System

The synthetic DES is a representation of an urban DES. The synthetic DES

was designed to allow flexibility to change and compare parameters such as weather,

energy equipment, and utility pricing. When first defining the synthetic DES, it was

important to understand what types of buildings typically are included in a DES.

Table 1 in Chapter 1 offers insight into this. From this table, nine different building

types were selected; these include:

•Public assembly

•Food service

•Food Sales

•Lodging

•Residential

•Workspace

•Mercantile/services

•Education

•Healthcare

Next, it was important to understand the energy load profiles for each of these

building types. The U.S. Energy Information Administration (EIA) conducts a

Commercial

Buildings Energy Consumption Survey (CBECS) and a Residential Energy

Consumption Survey (RECS) every six years. These surveys of the U.S. building

stock provide information on energy sources and uses, energy equipment, building

structural characteristics, and energy costs. The data collected in these surveys is

used for energy benchmarking, modeling, forecasting, and polivy formulation.

The U.S. Department of Energy developed commercial reference buildings in

partnership with the National Renewable Energy Laboratory, Lawrence Berkeley

National Laboratory, and Pacific Northwest National Laboratory. The commercial

reference buildings use the CBECS and the RECS as a starting point to develop

building energy models of sixteen building types in sixteen climate zones. From these

initial models, the national labs have further developed and calibrated the models to

fully represent a building; this includes HVAC equipment and specifications, thermal

zones, lighting, occupancy levels, and the thermal properties of the building

materials. These fully defined models can then be simulated in EnergyPlus™ and

produce results such as the energy load profiles of a building over a year (heating,

cooling, and electrical), the distribution of each load in how it is generated and used,

and the amount of natural gas the building consumes.

From the commercial reference buildings, thirteen were selected to be used in

the synthetic DES. Table 3 shows how the commercial reference buildings align with

the building types selected. This table also shows the number of buildings selected to

be included in the synthetic case. Additionally, this table includes the number of

floors and total square footage of each of the building types as defined by the

commercial reference buildings. It should be noted that the commercial reference

buildings selected are in climate zone 4A and that they were constructed in or after

1980 [46].

Table 3: Overview of buildings selected for the synthetic DES.

Building Type

Commercial

Reference

Building

Number of

Buildings

Number of

Floors per

Building

Total

Square

Footage

Food Service

Quick Service

Restaurant 10 1 25,000

Full Service

Restaurant 10 1 55,000

Food Sales Supermarket 2 1 90,000

Lodging Large Hotel 2 6 244,240

Residential Midrise

Apartments 20 4 674,800

Workspace

Large Office 7 12 3,490,116

Medium Office 5 3 268,140

Small Office 4 1 22,000

Mercantile/

Services

Stand Alone

Retail 12 1 299,544

Education Primary School 1 1 73,960

Secondary School 1 1 210,887

Health Care

Hospital 1 5 241,351

Outpatient Care 1 3 40,946

Public Assembly Park7 N/A N/A 400,000

7 It should be noted that the park is not a commercial reference building and that it does not consume

electricity, but acts as a public assembly space.

The layout of the synthetic DES can be seen in Figure 7. The buildings were

simulated in EnergyPlus™, and a picture of one of the simulated buildings and a

breakdown of its energy end uses can be seen in Figure 8. The yearly simulated

energy loads for the

synthetic case can be seen in Table 4 and Figures 9, 10, and 11, and the full data set

can be found in the supplemental Excel file.

Figure 7. Illustration of the synthetic DES.

Figure 8. Simulated hospital building energy model with break down of energy end

uses.

Table 4. Synthetic DES energy loads.

Buildings Yearly Electricity

Load (kWh)

Yearly Heating

Load (kWh)

Yearly Cooling

Load (kWh)

10 Quick Service

Restaurant 3,247,087 8,908 650,245

10 Full Service

Restaurant 5,090,376 13,153 1,085,671

2 Supermarket 3,526,883 971,908 198,469

2 Large Hotel 3,840,612 66,929 2,903,427

20 Midrise

Apartments 4,467,750 4,199,975 1,968,264

7 Large Office 38,337,512 2,514,805 28,929,792

5 Medium Office 3,193,892 667,737 1,769,402

4 Small Office 259,976 36,831 84,585

12 Stand Alone

Retail 4,430,904 1,411,241 1,196,628

1 Primary School 987,143 45,260 417,882

1 Secondary School 2,004,977 127,937 1,269,577

1 Hospital 5,362,508 33,985 4,001,647

1 Outpatient Care 848,502 18,770 594,399

Total 75,598,123 10,117,440 45,069,988

Figure 9. The total yearly electricity load of the synthetic DES, broken down by

the maximum, average, and minimum megawatt values.

Figure 10. The total yearly heating load of the synthetic DES, broken down by the

maximum, average, and minimum megawatt values.

Figure 11. The total yearly cooling load of the synthetic DES, broken down by the

maximum, average, and minimum megawatt values.

The energy loads were also normalized by the total building square footage

and by the city block square footage that the building occupies, to quantify the

building’s energy footprint. The buildings that pull the most energy are the quick and

full-service restaurants, supermarkets, the hospital, hotels, midrise apartments, and

the large offices. Many of these buildings operate on a 24/7 schedule or near 24/7

schedule. These buildings also have high energy loads due to the services that they

provide, and for many of these buildings it would be difficult to significantly reduce

these energy loads.

Figure 12. Total synthetic district energy system electrical load normalized by the total

building square footage. The buildings with the largest normalized load, quick service

restaurants, full-service restaurants, and supermarkets.

Figure 13. Total synthetic district energy system electrical load normalized by the city

block square footage occupied by the building. The buildings with the largest

normalized load, quick service restaurants, large offices, and hospitals.

Figure 14. Total synthetic district energy system heating load normalized by the total

building square footage. The buildings with the largest normalized load, supermarkets,

midrise apartments, stand-alone retail.

Figure 15. Total synthetic district energy system heating load normalized by the city

block square footage occupied by the building. The buildings with the largest

normalized load, midrise apartments, large offices, and supermarkets.

Figure 16. Total synthetic district energy system cooling load normalized by the total

building square footage. The buildings with the largest normalized load, quick service

restaurants, full-service restaurants, and the hospital.

Figure 17. Total synthetic district energy system cooling load normalized by the city

block square footage occupied by the building. The buildings with the largest

normalized load, large offices, the hospital, and hotels.

The synthetic DES was designed with a CUP to support the buildings within

the DES. The CUP includes a natural gas combined heat and power plant (gas

turbine, duct burner, heat recovery steam generator, and steam turbine), boilers,

chillers, and cooling towers. The combined heat and power plant was sized to handle

the 70th percentile of the heating load. The boilers were designed to handle 1.25 times

the total heating load. The chillers with the cooling towers were designed to handle

1.25 times the total cooling load. All three of these design practices are recommended

and followed by industry so as not to overburden the energy technologies. Table 5

shows the specifications of each energy technology in synthetic CUP.

Table 5. Synthetic DES CUP equipment specifications

Equipment Manufacturer Number of Pieces

of Equipment

Total Rated

Capacity

Gas Turbine Solar Turbines 1 6.5 MW of

electricity

Duct Burner Solar Turbines 1

15.2 MMBtu/hr

(HHV) of exhaust

gas

Heat Recovery

Steam Generator

Rentech Boiler

Services 1 40,000 lbs/hr of

steam

Steam Turbine Dresser Rand 1 3.5 MW of

electricity

Boilers Cleaver Brooks 2 40,000 lbs/hr of

steam each

Chillers Trane 4

1,600 tons of

chilled water per

chiller

Cooling Towers Baltimore Aircoil

Company 4

6,900 gpm of

chilled water per

cooling tower

3.2 George Washington University District Energy System

The George Washington University has a diverse DES. The university has a

7.4 MW combined heat and power plant that supports five buildings. 52 MW of off-

campus solar, which provides 50% of GWU’s electricity. Additionally, GWU has 497

kW of oncampus rooftop solar, and four solar water heaters that supply hot water to 4

buildings.

Finally, GWU has electric vehicle charging stations in two of its parking garages. A

map of the GWU DES can be seen in Figure 18, as well as pictures of on-campus

solar and solar hot water heaters.

Figure 18. The George Washington University district energy system. Including solar

water heaters (top left), and on campus rooftop solar (middle and bottom left).

While GWU has a diverse district energy system, the case study will only

look at the buildings supported by GWU’s Central Utility Plant (CUP). This section

of GWU’s full DES was chosen because it has the most recorded energy data and

supports key university infrastructure through a distributed energy network. The CUP

is in the basement of Ross Hall, the university’s medical research building. The CUP

supports five buildings with electricity, heating, and cooling, as seen in Figure 19.

Table 6 shows how the CUP supports each of the five building’s energy loads. The

CUP has a combined heat and power plant and boilers, chillers, and cooling towers to

support the energy loads of the five buildings. Figure 20 is an overview of the CUP’s

equipment, and Table 7 shows the specifications of the CUP’s equipment.

Figure 19. Buildings supported by The George Washington Central Utility Plant. Top

row, left to right, the Science and Engineering Hall, Ross Hall. Bottom row, left to

right. Fullbright Hall, JBKO Hall, Munson Hall.

Table 6. Building and energy loads supported by GWU’s CUP.

Name of the

Building Type of Building Electricity Heating Cooling

Ross Hall Academic and

Research

Supplied by the CUP,

supplemented by the

utility

Steam from

the CUP

Chilled water

from the CUP

Science and

Engineering Hall

Academic and

Research

Supplied by the CUP,

supplemented by the

utility

Hot water

from the CUP

Chilled water

from the CUP

JBKO Hall Residential Supplied by the utility Hot water

from the CUP

Chilled water

from the CUP

Munson Hall Residential Supplied by the utility Hot water

from the CUP

Chilled water

from the CUP

Fullbright Hall Residential Supplied by the utility Hot water

from the CUP

Chilled water

from the CUP

Figure 20. Overview of the equipment in the GWU CUP.

Table 7. GWU CUP equipment specifications.

Equipment Manufacturer Number of Pieces

of Equipment

Total Rated

Capacity

Gas Turbine Solar Turbines,

model Centaur 50 1

4.6 MW of

electricity and

52.94 MMBtu/hr

(HHV) of exhaust

gas

Duct Burner Coen, model 12-31

DBR 1

15.2 MMBtu/hr

(HHV) of exhaust

gas

Heat Recovery

Steam Generator

Rentech Boiler

Services 1 40,000 lbs/hr of

steam

Steam Turbine Dresser Rand 1

2.824 MW of

electricity at

40,000 lbs of steam

Boilers Cleaver Brooks,

model D-68E 2 40,000 lbs/hr of

steam per boiler

Chillers Trane 3 2,500 tons of

cooling per chiller

Cooling Towers Baltimore Aircoil

Company 4 6,690 gpm per

cooling tower

The CUP has many challenges, which are outlined in Figure 21. Some of

these challenges such as the seasonal ones, impact how the plant operates. In the

winter, when GWU is not supporting large cooling loads, they have excess electricity.

This excess electricity currently can’t support anything, and it cannot be sold back to

the grid, which is a challenge. The challenge results in the plant not operating the gas

turbine nor the heating recovery steam generator, and instead using the steam

generated by the boilers to support the steam turbine and the buildings. In the

summer the opposite problem occurs, there is an excess of steam because there is

minimal/no heating load. Currently, GWU has no options to use this steam, and due

to D.C. environmental regulations, GWU cannot release the excess steam into the

city. Therefore, GWU cannot have any excess steam, resulting in them again only

using their boilers to support the steam turbine and buildings. Additionally, when

there are no heating loads, the CUP stops operating, and the buildings are fully

supported by the utility. These challenges, which affect their operational ability, result

in decreased plant efficiency, losses in revenue, and an inability to support our own

buildings in an emergency.

Other challenges, such as GWU’s location impact how the CUP can be improved.

The CUP currently sits in a small basement, where there is no room for new

equipment.

Equipment that could help reduce the plant’s excess electricity and steam challenges.

Additionally, all the space around the CUP is developed and is prime D.C. real estate,

so expanding beyond our plant gets ruled out. Finally, because of our proximity to the

White House and State Department, we are restricted on the types of equipment that

we can operate. For example, the plant would not be allowed to switch the natural

gas CHP to hydrogen because there would need to be a hydrogen storage tank on

campus, and a storage tank of that size would pose an explosive risk. Finally, the lack

of metering, historical data, and data integrity at GWU immensely impacts how the

CUP can be analyzed and how future energy improvements can be planned.

Specifically, the lack of steam and chilled water BTU meters means that the yearly

heating and cooling loads are unknown for the buildings.

To study GWU CUP DES, the electrical, heating, and cooling loads for the

selected buildings needed to be determined. The hourly electrical load data for

Munson, Fulbright, and JBKO came from the Pepco CEO portal. The timeframe of

the data is August 13, 2019 - August 12, 2020. This was selected as the starting data

as it is the first recorded Pepco data set. The analysis could be further examined if the

data time frame extended from August 12, 2020, to the present, which could exclude

any abnormalities that the pandemic caused in the normal electricity use on campus.

The hourly electrical load data for Ross and SEH came from the WinMP portal

(Combined Pepco and Cogen total). This data set used the same timeframe to keep

consistency across the study. GWU does not currently meter heating or cooling use in

any of its buildings, and the few meters that capture this data for the CUP were

determined to be inaccurate. This determination came from examining the steam

Figure 21. Challenges of the GWU CUP.

•Winter: excess electricity

•Summer: excess steam

•Lack of metering

•Data integrity

•Lack of historical data

•700 kW buffer from utility

•Unable to sell excess electricty to the utility

•Environmental and noise regulations

•Limited space

•CHP operates at 57% of the rated capacity

•CHP electricity generation efficiency is 26%

•Unable to support surrounding GWU

buildings and the community in an

emergency

BTU meters, which were either reading zero, the same repeated value, or a value

inconsistent with the size of our steam system equipment. Therefore, the heating and

cooling loads for the selected buildings needed to be calculated.

The heating and cooling loads for the buildings supported by the CUP were

determined using a Python code due to a lack of building information; however, more

accurate models could be generated using OpenStudio once the information (HVAC

loop diagrams, building schedules, thermal zone layouts, floorplans, and building

material specifications) has been determined/collected. The calculations in the code

use assumptions based on ASHRAE’s Air Conditioning Contractors of America

Manual J, eighth edition (ACCA MJ8). The assumptions are as follows:

•Heating and cooling loads are split for the year (this is how GWU

currently operates, the exact day that the heating and cooling is flipped

on/off is based on the weather patterns of each individual year)

▪Heating months: November, December, January, February, March,

and

April

▪Cooling months: May, June, July, August, September, and October

•Schedules for people, lighting, and plug loads were not considered

•No gains or losses from equipment in the buildings (sensible or latent)

•No gains or losses from infiltration (sensible or latent)

•No gains or losses from the space for ductwork or crawl space (sensible or

latent)

•No gains or losses between zones or floors (sensible or latent)

•No gains or losses from ventilation or ductwork, no reheat (sensible or

latent)

•No gains or losses from radiation

•Room temperature for the buildings is a constant 70°F, no variations in

temperature supply to the rooms or variation between seasons

•Heat transfer rate for people = 255 BTU/hr/person (light seated work)

•Heat transfer rate for light bulbs = 27.28 BTU/hr/lightbulb

▪This value is for an 8 W LED light bulb

▪Excludes lighting use factors & ballast allowances

▪Cooling Load Factor = 1

•Heat transfer rate for plug loads = 512 BTU/hr/outlets

•Heat transfer coefficients [U]:

▪windows = 0.7 (double pane)

▪walls = 0.4 (cement blocks)

▪floor/ceiling = 0.2 (concrete)

•Latent cooling will not be extensively calculated but will be assumed at

20% of the total load

•Weather data was from Arlington Ronald Reagan Washington National

Airport and was taken from EnergyPlus™ (see chapter 3 section 3 for more

details)

Additional assumptions needed to be made about the five CUP buildings. These

assumptions are based on information gathered from inside the buildings and data

supplied by the university. The building assumptions are below. Additionally, it was

assumed that the average area of the windows on all the buildings was 6 sqft.

Ross Hall:

•Number of floors = 9 (two basement levels)

• Area of each floor = 45164.78 sqft

•Area of walls = 36540 sqft (east and west), 34020 sqft (north and south)

•Number of windows per wall o North: 0 basement level, 40% for levels 1-7

o East: 0 basement level or levels 1 and 2, 40% for levels 3-7 o South: 0

basement level or levels 1 and 2, 40% for levels 3-7 o West: 0 basement

level, 60% of walls on the rest of the levels

•Number of people in building = 1404 (677 faculty/staff, 712 students, 15

CUP

staff)

•Numbers of lights in the building = 10162 (2.5 lightbulbs per 100 sqft)

•Number of plug loads in the building = 28454 outlets (7 outlets per 100

sqft)

Science and Engineering Hall:

•Number of floors = 13 (five basement levels)

• Area of each floor = 61231.63 sqft

•Area of walls = 57330 sqft (east and west), 51870 sqft (north and south)

•Number of windows per wall o North: 0 basement level, 85% of wall on

each other level o East: 0 basement level, 85% of wall on each other level o

South: 0 basement level, 85% of wall on each other level o West: 0 basement

level, 85% of wall on each other level

•Number of people in building = 950 (100 people on levels B1-8, and 50

people on B2, no people in parking garages)

•Numbers of lights in the building = 16,839 (2.5 lightbulbs per 100 sqft)

•Number of plug loads in the building = 47148 outlets (7 outlets per 100

sqft)

Munson Hall:

•Number of floors = 9 (one basement level)

• Area of each floor = 5518 sqft

•Area of walls = 11340 sqft (east and west), 75 sqft (north and south)

•Number of windows per wall o North: 5 basement level, 16 on each other

level o East: 0 basement level, 5 on each other level o South: 5 basement

level, 16 on each other level o West: 0 basement level, 5 on each other level

•Number of people in building = 138 (number of student beds)

•Numbers of lights in the building = 1080 (15 rooms per floor, 8 lightbulbs

in a 2 person room)

•Number of plug loads in the building = 3476 outlets (7 outlets per 100 sqft)

Jacqueline Bouvier Kennedy Onassis Hall

•Number of floors = 9 (one basement level)

• Area of each floor = 9009.11 sqft

•Area of walls = 15750 sqft (east and west), 10710 sqft (north and south)

•Number of windows per wall o North: 3 basement level, 18 on each other

level o East: 10 basement level, 27 on each other level o South: 3 basement

level, 18 on each other level o West: 10 basement level, 27 on each other

level

•Number of people in building = 239 (number of student beds)

•Numbers of lights in the building = 2,736 (38 rooms per floor, 8 lightbulbs

in a

2 person room)

•Number of plug loads in the building = 5675 outlets (7 outlets per 100 sqft)

Fulbright:

•Number of floors = 9 (one basement level)

• Area of each floor = 5588.44 sqft

•Area of walls = 10,584 sqft (east and west), 8568 sqft (north

and south)

•Number of windows per wall o North: 4 basement level, 10 on

each other level o East: 4 basement level, 13 on each other level

o South: 0 basement level, 9 on each other level o West: 12

basement level, 17 on each other level

• Number of people in the building = 154 (number of student beds)

•Numbers of lights in the building = 1368 (19 rooms per floor, 8 lightbulbs

in a 2 person room)

•Number of plug loads in the building = 3521 outlets (7 outlets per 100 sqft)

To calculate the heating and cooling loads, the following equations were used:

Sensible heat addition from people, lighting, and plug loads:

Q people = (heat transfer rate of people) x (number of people)

Q lighting = (heat transfer rate of lighting) x (number of lightbulbs)

Q plug loads = (heat transfer rate of plug loads) x (number of outlets)

Q sensible = Q people + Q lighting

Latent heat addition from building conduction:

Q windows in the summer = (heat transfer coefficient of windows) x (number of windows) x

(area of

a window) x (dry bulb temperature – temperature setpoint) x

(latent cooling factor)

Q walls in the summer = (heat transfer coefficient of walls) x (total area of a walls – total

area of

windows) x (dry bulb temperature – temperature setpoint) x (latent

cooling factor)

Q roofs in the summer = (heat transfer coefficient of roofs) x (area of the roof) x (dry

bulb temperature – temperature setpoint) x (latent cooling

factor)

Q windows in the winter = (heat transfer coefficient of windows) x (number of windows)

(area of a

window) x (temperature setpoint - dry bulb temperature)

Q walls in the winter = (heat transfer coefficient of walls) x (total area of a walls – total area

windows) x (temperature setpoint - dry bulb temperature)

Q roofs in the winter = (heat transfer coefficient of roofs) x (area of the roof) x (temperature

setpoint - dry bulb temperature)

Q latent summer = Q windows in the summer + Q walls in the summer + Q roofs in the summer

Q latent winter = Q windows in the winter + Q walls in the winter + Q roofs in the winter

Heating and cooling loads:

Q heating = Q sensible + Q latent winter

Q cooling = Q sensible + Q latent summer

The hourly heating and cooling loads were calculated over one year, using the

above calculations, and the results of these calculations can be seen in Figures 22 and

23. The results of the heating and cooling calculations were verified against the size

of the CUP equipment. The hourly electrical loads for the five buildings can be seen

in Figure 24.

The full data set of the GWU CUP energy loads can be requested.

Figure 22. GWU CUP DES heating load, with maximum, average, and minimum

values for each month.

0.0

5.0

10.0

15.0

20.0

November December January Febuary March April

Heating (MW)

GWU CUP DES Heating Load

Max Average Min

Figure 23. GWU CUP DES heating load, with maximum, average, and minimum

values for each month.

Figure 24. Yearly GWU CUP DES electrical load, with maximum, average, and

minimum values for each month.

0.0

5.0

10.0

15.0

20.0

May June July August September October

Cooling (MW)

GWU CUP DES Cooling Load

Max Average Min

It can be seen from Figure 24 that the CHP system, which was designed to

meet the electrical load requirements of Ross Hall and the Science and Engineering

Hall, is overdesigned: the maximum total load of all five buildings (5.8 MW) falls

under the CHP rated capacity (7.42 MW). GWU CUP has excess capacity that could

supply the three dorms with electricity as well as other campus buildings.

3.3 Weather

The weather data used for both case studies was for Arlington Ronald Reagan

Washington National Airport and was taken from EnergyPlus™, which is

representative of the Mid-Atlantic weather [47]. This weather file is a typical

meteorological year (TMY) and is an average of 30 years of weather data. This

weather file provides 68 different weather measurements for an entire year, including

wet and dry bulb temperature, solar irradiance, illuminance, and sky cover [48].

3.4 Utility Pricing

The utility pricing for both case studies follow Pepco schedules for electricity

and Washington Gas schedules for natural gas. Pepco and Washington Gas are the

utilities for the Washington, D.C., area. All Pepco schedules follow the rules around

billing and rating periods, as seen in Figure 25. The energy charges vary by season

(June – October and November – May) and time (on-peak, off-peak, and intermediate

peak). The demand charge is applied to the max kW value during the on-peak time or

to the maximum kW value across all three times, whichever is highest. The study will

use two specific Pepco schedules, GT-3A and MGT-LV. The first is Pepco GT-3A,

which is used for commercial buildings with power generation. The second is Pepco

MGT-LV, which is used for large commercial buildings with no power generation. A

breakdown of the fees and costs of each of these two schedules can be seen in Figure

26. Both rate structures match the rates seen on GWU utility bills for the five CUP

buildings.

Figure 25. Pepco billing and rate guidelines [49].

Figure 26. Pepco rate structures used in the case studies. Pepco GT3A for commercial

buildings with power generation, and Pepco MGT-LV for commercial buildings with

no power generation [50, 51].

Washington Gas similarly has three charges within all rate structures: flat

charge, energy charge, and demand charge. The demand charge is applied to the

maximum MMBtu/hr in the on-peak months (November – April). It should be noted

that many Washington Gas rates are contractually negotiated, so there can be

variations between the set schedules and how buildings are billed. This is the case for

GWU, where not all the utility bills match the rate structure that GWU follows. The

case studies use two Washington gas rate structures. The first is rate 7, which is used

for commercial buildings with power generation. The second is rate 2, which is used

for commercial buildings with no power generation. A breakdown of the fees and

costs of each of these two schedules can be seen in Figure 27.

Figure 27. Washington Gas rate structures used in the case studies. Rate 7 for

commercial buildings with power generation, and Rate 2 for commercial buildings

with no power generation [52]. GWU rate seen on utility bills, which differs from the

set rate structures.

Chapter 4: District Energy System Case Study Analysis

This chapter will demonstrate a method to plan and analyze a DES in an

urban area using a combination of BEM tools. This chapter will use the work

discussed in Chapter 3 as the inputs. Then, the chapter will use some of the BEM

tools discussed in Chapter 2 and add a few others for the analysis. Lastly, the chapter

will explore the direct outputs of the analysis and set up Chapter 5 to discuss the

calculated outputs. An overview of this analytical flow is seen in Figure 28. Various

tools will be used to analyze each technology separately, and then the results will

come together in Excel for further analysis and calculations. This method was

developed because it offers the most component flexibility for each energy

technology. Also, as previously discussed, not all energy modeling tools can model

every energy technology. Each of the tools used in this method are used by the energy

industry, and most have been either developed by or for the DOE National

Laboratories. This method allows communities to plan for and analyze the most

common energy generation and storage technologies currently used in DESs: solar

panels, battery storage, ice storage, hot and chilled water storage, boilers, chillers,

cooling towers, gas turbines, heat recovery steam generators, duct burners, and steam

turbines. However, more tools should be added to this methodology as they are

developed to account for new trends in DESs: hydrogen and renewable fuels,

geothermal, biomass, and fuel cells.

Figure 28. Overview of the analysis of the DES case studies.

4.1 Solar Panel Analysis

This study used a commercially available solar PV panel to analyze the

impact solar generation would have on each of the buildings in the DES case studies.

The specifications of the solar panel can be seen in Table 8. The solar panels were

modeled in OpenModelica®. The model used the specifications discussed in Table 8

and the weather file to output yearly electricity production. The model assumed that

there was no shading on the panels from trees or other buildings, that the panels were

flat on the roof, and that there was no limitation on electricity production. The

modeled maximum generation from the solar panels was 161 W, which falls below

the rated nominal power of 270 W. One solar panel was modeled for this study, and

the results were multiplied by the calculated number of solar panels on each

building’s roof. The modeled electricity generation from one solar panel for the first

seven days of the year can be seen in Figure 29. The full yearly solar output data set

can be found in the supplemental file.

Table 8. Solar PV Panel Specifications [52].

Manufacturer Type Nominal

Power

Maximum

Power

Voltage

Rated

Efficiency

Dimensions

AEG Polycrystalline 270 W 31 V 16.6 % 3 x 5 feet

Figure 29. Modeled electricity production from one solar panel.

The solar panels were only applied to 40% of the building’s roof because

urban buildings typically have mechanical and electrical equipment (HVAC) on the

roofs, and some urban roofs have dedicated spaces for occupants (lounge areas,

pools). Solar panels were applied to all of the buildings in the synthetic case. In the

GWU CUP DES case, solar panels were only applied to residential buildings, as Ross

Hall and the Science and

Engineering Hall have their roofs fully occupied with mechanical and electrical

equipment. The breakdown of the number of solar panels and the modeled energy

generation per building can be seen in Table 9. The full data set of the yearly

generation per building can be found in the supplemental file.

Table 9. Electricity generation from rooftop solar panels.

Type of Building 40% of the

rooftop area [m2]

Number of Solar

Panels

Total Generation of Electricity

per Building (kWh/year)

7 Large Office Buildings 10,808.9 7,506 1,770,931

5 Medium Office Buildings 3,321.5 2,306 544,067

4 Small Office Buildings 817.6 567 133,775

10 Full Service Restaurants 2,043.9 1,419 334,792

1 Hospital 1,793.8 1,245 293,740

2 Large Hotels 1,512.7 1,050 247,732

20 Midrise Apartments 6,269 4,353 1,027,027

1 Outpatient Care 507.2 352 83,049

1 Primary School 2,748.4 1,908 450,165

10 Quick Service Restaurants 929 645 152,178

1 Secondary School 3,918.4 2,721 641,980

12 Stand Alone Retail Stores 11,131.4 7,730 1,823,781

2 Supermarkets 3,344.5 2,322 547,842

Total 49,145.6 34, 124 8,051,060

Fulbright Hall 183.2 127 29,963

JBKO Hall 311.6 216 50,962

Munson Hall 190.8 132 31,143

Total 685.6 475 112,068

Table 9 showcases how important it is to consider the energy density of the

technologies that are being used when analyzing an urban DES. This point is

highlighted by the synthetic case, where the maximum generation capacity of all the

solar panels in the case (34,124 panels) is 5.5 MW and takes up the equivalent of 9.2

football fields, whereas the combined heat and power plant can supply 8.9 MW in the

equivalent of 1 football field.

Other metrics are also important to consider: the percentage of the building’s

electrical load that solar can provide and the reduction in the annual electricity bill.

The results of these metric calculations can be seen in Table 10. These calculations

highlight solar’s varied impact on a building; the values ranged from 5-51%. This is

seen especially in buildings with high energy intensity loads, like the hospital, where

solar could only support 5% of the building’s electrical load. The hospital, a critical

piece of infrastructure, would benefit the most from having energy generation

technologies that can support its loads in an emergency.

Table 10. Impact of solar generation on electricity use.

Type of Building

Percentage of the Building’s

Electrical Load that Solar can

Provide

Reduction of Energy Bill

from Solar Generation

7 Large Office Buildings 5% 3%

5 Medium Office

Buildings 17% 11%

4 Small Office Buildings 51% 35%

10 Full Service

Restaurants 7% 9%

1 Hospital 5% 3%

2 Large Hotels 6% 4%

20 Midrise Apartments 23% 16%

1 Outpatient Care 10% 5%

1 Primary School 46% 37%

10 Quick Service

Restaurants 5% 7%

1 Secondary School 32% 21%

12 Stand Alone Retail

Stores 41% 26%

2 Supermarkets 16% 12%

Average 11% 8%

Fulbright Hall 17% 15%

JBKO Hall 18% 16%

Munson Hall 24% 20%

Average 20% 17%

4.2 Battery Storage Analysis

When analyzing battery storage for a DES, it is important to choose a battery

size that will reduce the highest percentage of peaks (flatten the demand profile)

while also being mindful of the size and the cost. It is important to flatten the demand

profile because this will reduce the amount of energy being used during expensive

on-demand times. The more peaks a battery can flatten the quicker the return on

investment will be. Selecting a commercially available battery storage technology

and using industry specifications are also important. This is crucial because

communities need technologies that will last (and have been proven to do so) to make

their return on investments and for the batteries to be serviced for their lifetime.

Table 11 shows the specifications of a commercially available battery storage

technology. While multiple batteries can be used to increase the rated power, a

singular battery (979 kW) takes up 28.9 ft x 5.4 ft x 9.1 ft. This is a considerable

amount of space for most urban buildings. For example, in the GWU CUP DES case,

there are only two options, options 2 and 10, that would fit a battery of this size.

However, both options could accommodate up to four of these batteries, providing a

total of 3.9 MW. Both options are parking garages where parking spaces would need

to be given up to install the batteries, and this may be a conflicting issue for an

organization. The options for battery storage locations for the GWU CUP DES can be

seen in Table 12.

Table 11. Battery storage specifications [54].

Manufacture

Type Rated

Powe

Rated

Energ

Interconnectio

Roundtri

Efficienc

Dimension

Tesla

MegaPack

Lithiu

m Ion

979

3,916

kWh

480 V

AC 3 phase

93.7% 28.9 x 5.4 x

9.1 (feet)

Table 12. Battery storage location options for the GWU CUP DES.

Space Type Size (feet) Location

Option 1 Mechanical Room 8.9 x 14.8 School of Engineering

and Applied Science

Option 2

Parking Garage

Adjacent to a

Mechanical Room

120 x 70 (half of

the parking

garage)

School of Engineering

and Applied Science

Option 3 Boiler Room 22.6 x 12.7 JBKO, Residential

Dorm

Option 4 Electrical Room 2 x 4 JBKO, Residential

Dorm

Option 5 Storage Room 3 x 10 JBKO, Residential

Dorm

Option 6 Storage Room 11 x 6 Fulbright, Residential

Dorm

Option 7 Storage Room 26.7 x 14.8 Fulbright, Residential

Dorm

Option 8 Chiller Room 13 x 5 Ross Hall, Houses the

Central Utility Plant

Option 9 Chiller Pump Room 18 x 16 Ross Hall, Houses the

Central Utility Plant

Option 10

Parking Garage

Adjacent to the Central

Utility Plant

135 x 36 Ross Hall, Houses the

Central Utility Plant

Option 11 Parking Garage

Adjacent to the Central

Utility Plant

54 x 9 Ross Hall, Houses the

Central Utility Plant

MATLAB’s® Simulink® BESS model was used to analyze battery storage options for

the DES cases [55]. The battery has a simple design with three blocks, an electrical

demand that feeds into the battery and a controller that then supports a system. In

this case, the electrical demand is the yearly synthetic DES electrical load. The

battery inputs are rated power, power conversion efficiency, nominal battery capacity,

initial state of charge, system auxiliary power, and battery charge/discharge level.

The values used for the inputs in the synthetic case can be seen in Table 13. The

controller, which tells the battery when to charge and discharge, was set at 10 MW

for this case. The model outputs, which shows how the battery supports the system,

are state of charge and change in electrical demand, which can be seen in Figures 30

and 31, respectively.

The value for the controller was selected because it is the rated generation

capacity of the CUP, and when you are operating a CHP (the power source in the

synthetic CUP), it is important to operate it at its rated capacity for optimal

efficiency. Adding a battery to a CHP with a fluctuating load, allows for more use out

of the CHP and for more optimal use, which helps in making a return on investment

on the CHP system. This concept can be seen in Figure 31, where the gray line is the

rated capacity of the CHP (where it would ideally operate 24/7). Here, the blue line

shows how the CUP meets demand without the battery, and the orange line shows

how it meets demand with the battery. The results of using a battery show significant

reductions in the peaks of meeting demand, which means the CHP is operating closer

to its ideal capacity. This also means reductions in electrical peaks that must be

supported by the utility (electrical demand that extends beyond the ability of the CHP

and battery to meet). This means lower electricity bills and aids in making the return

on investment on the battery. Ideally, you could get a large enough battery such that

the CHP always operates at its ideal capacity; however, when planning for a battery

the planner must consider the size and cost of the battery. Therefore, a balance must

be achieved between maximizing support to a system and minimizing cost while

fitting into the available space.

Table 13. Battery model input parameters.

Rated power 4.8 MW

Power Conversion Efficiency 90%

Nominal battery Capacity 50,000 kAh

Initial State of Charge 0%

System Auxiliary Power 5%

Charge battery when consumption is below 10 MW

Discharge battery when consumption is above 10 MW

Figure 30. State of charge of the 4.8 MW battery used in the synthetic DES case.

Figure 31. The effect of battery storage on the electrical demand of the synthetic

DES case.

4.3 Additional Tools

This section will include a discussion of additional tools included in the

methodology. There will be no results in this section as the results from using each of

these tools could not be verified. However, these are very strong tools that are used

by industry professionals. Additionally, these tools were all developed by or for the

National Laboratories. EnergyPlus™ and OpenStudio® were used to model thermal

storage (ice storage, hot and chilled water tanks) and the chilled water loops (chillers

and cooling towers); however, these tools were discussed in section 2.2.1 and will not

be included here.

4.3.1 Combined Heat and Power Plants

Combined heat and power plants (CHP) have many pieces of equipment that

work together and have many interdependencies between each piece of equipment.

Therefore, when analyzing a CHP for a DES it is important to model as many of the

components and specifications as possible to capture the details of the plant. The

Solar Turbines Engine

equipment (gas turbine, duct burner, heat recovery steam generator, boiler) except for

the steam turbine. The STEPP modeling tool is used in NREL’s REopt™ tool and can

be seen in Figure 32.

•Inputs

•Site Conditions: ambient temperature, humidity, pressure, and

elevation, natural gas supply pressure

•Equipment: size of gas turbine, duct burner use

•Return Conditions: steam or electricity demand, steam pressure and

temperature

•Outputs

•Gas Turbine Conditions: exhaust gas temperature, mass and

volumetric flowrate, and emissions, natural gas use, power produced,

efficiency, and compressor power

•HRSG Conditions: steam mass flowrate, temperature, and efficiency,

duct burner firing temperature, duct burner fuel use, exhaust gas mass

flowrate, and temperature, feedwater mass flowrate and temperature

•Limitations:

•Limited turbine sizes, only natural gas, no steam turbine, no

economizer, can’t specify boiler size, assumes plant outages

4.3.2 Boilers and Steam Turbines

The DOE Steam System Modeler Tool (SSMT) was used to model the boilers

and steam turbine. This simple tool allows industry professionals to quickly calculate

values when working in the field. The tool utilizes steam tables and simple

thermodynamic equations. It should be noted that DOE has now integrated this tool

into a larger calculation platform called MEASUR, so SSMT is no longer available to

be downloaded.

The SSMT has the following assumptions, inputs, outputs, and limitations:

•Assumptions:

•Deaerator provides feedwater near the boiling temperature for the

deaerator's set operating pressure

•Steam, Boiler, and Blowdown Pressure are the same

•Combustion Efficiency is the % of fuel energy that is directly added to

the feedwater and not otherwise lost or used

•Blowdown Rate is the % of incoming feedwater mass flow rate that

leaves the boiler as a saturated liquid at boiler pressure

•The model does not include any evaluation of any complex boiler

configurations or fluctuations in operation

•Energy from motors (pumps, fans, etc.) are not considered

•Inputs

•Pressure, temperature, specific enthalpy, specific entropy, saturated

quality, combustion efficiency (%), generator efficiency (%),

isentropic efficiency (%), blowdown rate (%), deaerator vent rate (%),

heat loss (%), condensate return (%), steam mass flow, feedwater mass

flow, initial hp steam usage, initial hp energy usage

•Outputs

•Fuel usage, power production, steam production

•Limitations:

•Cannot not calculate values over a period of time, cannot select a

specific type of boiler

To use this tool to model year long scenarios, the tool was combined with python.

The python model ran a loop of the SSMT calculations, and outputted yearlong

results.

Chapter 5: Improving the Impact of Solar Generation on Urban Hospitals

This chapter will use the techniques and methods discussed in the past

chapters to analyze a hospital in the synthetic urban DES. The analysis explores how

impact of solar generation and battery storage on urban hospitals.

5.1 Hospital Energy Loads

Hospitals must support critical loads in an emergency, and these loads are

substantially larger compared to the energy loads of other commercial buildings, as

seen in Figure 33. Figure 33 is from the DOE EERE BTO [56] and is similar to

Figure 34, which is the same graph applied to the synthetic DES. The larger energy

loads of the hospital mean that renewables such as rooftop solar will have a small

impact on offsetting the load.

Figure 33. Energy use per building for common commercial building types

[56].

Energy Use Per Building in the Synthetic DES Case

0 2000 4000 6000 8000 10000 12000

Energy Use (MWh)

Figure 34. Energy use per building in the synthetic DES case.

1 Quick Service Restaurant

Full Service Restaurant1

1 Supermarket

1 Large Hotel

Midrise Apartments1

1 Large Office

1 Medium Office

1 Small Office

1 Stand Alone Retail

Primary School1

1 Secondary School

Hospital1

Outpatient Care1

This analysis explored how urban hospitals, specifically in Washington D.C.,

can use solar generation and battery storage to offset the hospital’s energy loads,

improve its ability to withstand an emergency, and evaluate the resulting payback

period. Additionally, this analysis looked at using district solar, otherwise known as

community solar. Community solar for this study meant that electricity produced by

solar panels on all community buildings could be used solely to support the hospital.

This study only looked at retrofitting the hospital, so changes to the building

geometry, materials, thermal zones, and building codes will not be examined.

Additionally, it was assumed, and therefore not studied, that plug loads (such as

medical equipment), temperature setpoints, and operational schedules cannot be

changed. Finally, it was assumed that the hospital will be operating as it was

commissioned, and therefore, the impact of measurement and verification retrofits

was not studied. While this study was conducted in D.C., the analysis can be done for

hospitals and other buildings worldwide. This study is important because it can help

hospitals, especially those with limited resources and support, understand and

prioritize the most impactful changes they can make in their buildings to increase

reliability and decrease costs. At the same time, this study will also help hospitals to

reduce their strain on the grid and reduce carbon emissions. Reducing the electrical

strain and improving the reliability of one building strengthens the nation’s electrical

grid.

5.2 Technoeconomic Metrics

The study used the previously discussed synthetic case electrical load, solar

generation, utility pricing, weather, and equipment specification to examine the

impact of solar generation and battery storage on the hospital electrical load. The

study used one battery from the Table 11 specifications and the MATLAB® BESS

model. The results of the solar and battery model were imported into an Excel file.

From there, three metrics were calculated: percent offset of load, payback period, and

reliability. The first metric, percent offset of the load, quantifies the impact a

technology has on supporting a building’s

load and is described by Equation 1.

% offset = 𝐿𝑜𝑎𝑑 𝑤𝑖𝑡ℎ 𝐵𝑢𝑖𝑙𝑑𝑖𝑛𝑔 𝐶ℎ𝑎𝑛𝑔𝑒

*100 (Equation 1)

𝐵𝑎𝑠𝑒 𝐶𝑎𝑠𝑒 𝐿𝑜𝑎𝑑

The second metric used is payback period. This metric is used to understand

the impact of purchasing new technologies for the building changes versus the energy

saved by those changes and is described in Equation 2. Table 14 includes the capital

and operation and maintenance (O&M) costs of the new technologies. The O&M cost

for solar PV was not supplied by the manufacturer and was instead taken from EIA’s

Annual Energy Outlook 2022 [57]. The annual energy savings were calculated using

the Pepco utility rates described in Table 26. It was assumed that the cost of utilities

did not change as a function of year. Additionally, taxes, tax incentives, and any

federal/D.C. programs related to the technology changes were not considered. Also,

the ability to sell excess electricity to the grid was not considered. Finally, it was

assumed that the technologies’ performance did not degrade with time.

𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝐶𝑜𝑠𝑡+ 𝑌𝑒𝑎𝑟𝑙𝑦 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑜𝑛 𝑎𝑛𝑑

𝑀𝑎𝑖𝑛𝑡𝑒𝑛𝑐𝑒 𝐶𝑜𝑠𝑡

Payback Period =

𝐴𝑛𝑛𝑢𝑎𝑙 𝐸𝑛𝑒𝑟𝑔𝑦 𝑆𝑎𝑣𝑖𝑛𝑔𝑠

(Equation 2)

Table 14. Capital and O&M costs for new building technologies.

Building Technology Capital Cost

Yearly Operation

and Maintenance

Cost

Solar PV Panels for Hospital $610,050 $3,200

Solar PV Panels for Hospital &

Community $12,782,630 $67,056

Lithium Ion Battery $2,473,180 $8,440

The third metric used is reliability. According to DOE, reliability is defined as

“the ability of a system or its components to withstand instability, uncontrolled

events, cascading failures, or unanticipated loss of system components.” This study

only focused on one failure type, a two-hour electrical power outage. This was

selected from Pepco’s Reliability Metrics for the District of Columbia as an average

of CAIDI (customer average interruption duration index) metrics between 2015 and

2019. The two-hour power outage occurred on June 1st from 12 -2 pm. Reliability

was quantified as a loss of energy: energy demanded by the system that is not met.

5.3 Results

The results of the study can be seen in Table 15, and the data used can be seen

in the supplemental Excel file. The interpretation of the results may vary based on the

hospital’s energy, environmental, and cost goals; however, generally the best option is

to use community solar and battery. This option provides the best reliability, payback

period, and ability to offset the electrical load. This option requires 26,047 solar

panels and a 1 MW battery to support 97% of the electrical load. This offset does not

include the energy required for the natural gas usage of the building and highlights

how difficult it is to reach net zero emissions for an urban hospital using renewables.

The hospital’s decision should also weigh the ethical and environmental impacts of

manufacturing so many solar panels and a large battery.

The second-best option is to utilize only community solar, an option that still

ranks highly in all three metrics. Community solar is beneficial, especially in cases

like this one where solar does not adequately support each individual building. But,

when all the buildings come together, one important building in the community can

be fully supported. This option is also good if the hospital does not have the space for

the battery or is concerned about the fire risk of a battery. The third best option is to

utilize the hospital’s rooftop solar and a battery. This option is best if the hospital is

concerned about supporting small load fluctuations or power outages; however, it

would not be suitable to support the hospital over a multi-day failure event. The long

payback period of the third option may be acceptable to a hospital because any loss

of energy for even a minute can have catastrophic cost implications for the hospital.

Similarly, the adding a battery in this option or the first can be justified because in

emergencies, batteries start immediately, and backup generators can take minutes to

start up. Additionally, generators that have not been used for years risk not starting in

an emergency.

Table 15. Metric results for the study.

Scenario Type % Offset of

Electricity Load

Payback Period

(Years)

Reliability

(kWh)

Hospital 0 N/A 2,798

Hospital with Solar 2.7 3.1 2,471

Hospital with Solar and Battery 2.7 15.8 0

Hospital with Community Solar 88.2 5.1 0

Hospital with Community Solar

and Battery

97.3 2.6 0

This study examined two different building changes; however, there are many

more changes which could be studied. The potential changes are:

•Use of alternative energy storage (e.g., thermal storage)

•Use of community distributed heating and cooling

•Electrification of the hospital equipment

The study could also examine more failure events such as:

•Long duration power outage

•Extreme weather

•Physical Attack

•Equipment failure

Finally, this study could be furthered by comparing the results of this study to a case

study of a hospital that has implemented these changes or is planning to.

Additionally, the case study could be used as a point of comparison for how

measurement and verification retrofits compare to the impact of the studied building

changes.

Chapter 6: Conclusion

This study examined how to develop a methodology to plan DESs in urban

communities and answer important questions about achieving community climate

action goals through renewable initiatives. The study discussed current urban

planning methods and how those methods fail for DESs. Then, the study discussed a

new method that could be used until better software is developed. Next, two case

studies were created to study the new methodology. Finally, the methodology was

applied to the scenario of an urban

hospital.

This study provided insights into the challenges of planning DESs in urban

areas, current software, current building systems, and implementing climate action

goals. One theme that became apparent through this work was that every DES is

different in terms of the energy technologies they have and can support and the goals

that they are trying to achieve. Therefore, it is important when developing a master

plan to have a method that is general enough to be applied to every kind of DES and,

simultaneously, take in the specifics of each DES so as not to lose the unique

challenges they face. Communities will likely need a lot of support during this

planning and implementation process, as many do not have the required information

to do this work. Some of that support needs to be people who can help take

measurements and find building specifications, but there should also be a team

working on developing general specifications that could be used in the analysis in the

case that certain data cannot be obtained.

Another theme was that all DES face challenges, even those with the best

energy management teams supporting them. A lot of these challenges are challenges

that buildings in general face, such as a lack of metering, aging infrastructure,

regulatory compliance, costly upgrades, and delayed or skipped preventative

maintenance. While it will be beneficial to have better tools to analyze and model

solutions for these challenges, ultimately, there needs to be a shift in leadership. A lot

of organizations’ leadership has the mentality that if the building/systems are on, then

nothing needs to change; essentially, if it's not broken, don't fix it. However, to start

achieving the big climate action goals, the basic challenges at the building level need

to be addressed. These challenges will take a significant amount of money to fix.

Some of the fixes will pay for themselves through energy savings. There are some

programs that organizations can take part in, such as utility service performance

contracts or energy service performance contracts, but there will still be a need for

significant capital to make all the upgrades that have been deferred for years. Some

buildings have pushed old HVAC equipment 10-20 years over its expected lifetime

because there is no budget for new equipment, but running inefficient equipment

causes high utility bills. Therefore, leadership needs to recognize and support these

challenges, and that work starts with better communication. For good

communication, there needs to be a good energy support team in place for the DES.

Another theme that was seen and is a major challenge to DESs and buildings

is a shortage of trained working professionals. These include electricians, mechanics,

plant operators, energy managers, and engineering support staff. Some organizations

recognize the importance of having a strong energy support team, but they simply

cannot find people to fill the roles. This challenge is expected to worsen for the DES

and building energy community, as many of the current working professionals are

reaching retirement age, and not enough people are graduating to fill the roles. More

work must be done to recruit and properly train the next generation of workers in this

field.

Additionally, when most communities are making climate action plans, they

are not considering the state of the distribution lines (water, natural gas, and

electricity). Many distribution lines in the U.S. are old and are not rated to handle

more demand. This is especially seen for the electrical distribution lines. Many

renewable energy initiatives will require more power to be supplied across the lines.

The power lines will either need to be replaced with higher voltage limit cables (this

is costly and time-consuming) and new transformers, or the buildings will require

more current, which could result in blowing a transformer (this has been occurring

more frequently in the last 5-10 years). This challenge is also preventing some

renewable energy projects from happening, especially larger projects, including

utility-scale renewables and storage, and Power Purchase Agreements for solar and

wind farms. Finally, this challenge is also one that needs to be considered when

communities are thinking about reliability and resiliency measures because a system

is only as resilient and reliable as its distribution lines. For example, if a community

has a microgrid that can support that community in an emergency, but has above-

ground power lines, then if the community power lines get taken down by a storm,

the microgrid is useless. Therefore, when planning a DES or energy system for a

climate action goal, it is imperative to be mindful of the distribution lines, and to

account for the cost needed to upgrade these systems.

The last theme was that renewable energy initiatives may never be enough for

urban DESs to reach climate action goals. This was seen especially for the GWU

case, where battery and thermal storage options did not make sense due to the energy

loads being significantly lower than the equipment’s rated capacity, limited space,

and fire restrictions. For this case, expanding the current distribution network to

support more buildings would make more sense. GWU has the equipment capacity to

support more buildings with heating, cooling, and electricity. Additionally, GWU

could participate in utility incentive programs, like demand response. A demand

response program could allow GWU to shed load from the utility, use the CHP to

support it, and get paid by the utility to do so. Utility incentive programs are

becoming more popular with DESs and buildings to pay for resiliency measures (the

utility pays an organization that has generators or batteries that can be on standby to

shed load). In general, the utilities, states, and the government need to create more

programs and incentives to support urban communities where traditional climate

action strategies do not work.

While there are many challenges that the DES and building energy

communities face, there are many organizations that are working diligently on these

challenges. Therefore, the author is confident and excited to work in this industry and

to see what the future brings.

Urban Energy Master Planning Methodology for District Energy Systems

Students also viewed