Case study

SALMONES PUYUHUAPI (A)

UVA-QA-0749

Rev. Jan. 28, 2011

( 57 )

Osvaldo Correa, CEO of Salmones Puyuhuapi (SP), was in a tense meeting with bank

officials

negotiating

the terms

of his firm’s line of

credit when

he received a text on his

BlackBerry. The news was not good. Correa tried to remain calm in front of the bankers as he

read the terse messa e about a suspected outbreak of the infectious salmon anemia (ISA) virus in

a competitor’s salmon farm. Correa knew that if the virus were to spread to his firm’s Jacaf Fjord site in northern Patagonia, Chile, it could wipe out his nearly one million salmon two months before they were to be harvested. That would have a devastating impact on his firm’s cash flows. Excusing himself to go to “the services” (the only excuse he could come up with), Correa e-

mailed his assistant

to set up

a meeting

later that day with Jorge Richards, his operations

manager, and Sergio Rivas, the company’s veterinarian.

Salmon Farming in Chile

Chile was a relatively small country of nearly 16 million inhabitants whose economy was driven principally by mining, agribusiness, forestry, and aquaculture. In particular, the salmon industry, although relatively new, had been wildly successful. The industry was established in the early 1980s to take advantage of natural conditions in the south of the country (moderate sea

temperatures, sheltered sites, and ideal salinity levels). It achieved nearly 20 years of 42%

annualized growth, allowing it to join Norway as the world’s largest salmon producers. In 2006,

these two countries

shared 78% of global production. Salmon

farms, which thrived in the

southern regions of hile, dominated the local economies and led to high rates of employment.

In April 2007, however, the industry was changed forever when it was discovered that high fish mortality rates at a few Chilean sites owned and operated by Marine Harvest, a large Norwegian company, were due to the ISA virus. This virus was well known in the aquaculture industry because of the catastrophic effects it had had on the salmon industries of Canada, Great Britain, and Norway. ISA outbreaks sometimes wiped out entire million-fish farms in a matter of

months.

For smaller

firms, the

cash flow

consequences of these

outbreaks

were particularly

This case was prepared by Rafael Echeverria (MBA ’11) and Phillip E. Pfeifer, Richard S. Reynolds Professor of Business Administration. It was written as a basis for class discussion rather than to illustrate effective or ineffective handling of an administrative situation. It is based on general experience and should not be used as a source of primary research. All names have been disguised. Copyright  2010 by the University of Virginia Darden School

Foundation, Charlottesville, VA.

All rights

reserved.

To order

copies, send an e-mail to

[email protected]. No part of this publication may be reproduced, stored in a retrieval system,

used in a

spreadsheet,

or transmitted in any form or by any

means—electronic, mechanical, photocopying,

recording, or otherwise—without the permission of the Darden School. ◊

devastating because a significant portion of revenues from one harvest were used to start or continue the 30-month cycles at other sites.

To make matters worse, Marine Harvest had initially hid information about the presence of the virus. This lack of warning to other companies allowed the disease to spread to other farming sites and production zones1 because the uninformed companies took no precautions. In early 2007, the virus was spread inadvertently by third parties contracted to provide feed and harvesting services. Once the industry learned the ISA virus had made its way to Chile, a host of process changes were immediately implemented to prevent its spread. In addition, Chilean authorities quickly banned the transfer of nondiseased salmon from infected zones out of concern that transferring apparently healthy salmon could actually help spread the disease to healthy zones. Although this policy helped slow down the spread of the virus across zones, it was at the increased risk of infecting all sites in an infected zone.

After the 2007 outbreak had run its course, the improved processes and regulations reduced the spread among zones; however, the threat was not eliminated and occasional outbreaks did occur. The last such outbreak had been six months earlier.

Atlantic Salmon Production Cycle

In the wild, adult Atlantic salmon spawned in natural freshwater streams, where their eggs hatched and juveniles grew through several distinct stages. Eggs quickly hatched into “alevins” that remained in the breeding ground and fed from the remaining nutrients in their yolks. After an alevin absorbed its yolk sack, it became a “fry.” In this stage, the fish was an active swimmer and soon left its breeding ground in search of food. In its final freshwater stage, when it became physiologically ready to travel to seawater, a fry became a “smolt.” In the wild, the duration of these freshwater phases varied between one and five years.

During the seawater phase, which took another one to four years, salmon acclimatized to salinity and grew to adulthood in the open ocean. With continued time and growth, adult salmon became “grilse.” At the grilse phase, they were ready to return to their natal streams. (Salmon were well known for swimming upstream to spawn at their birthplace.) The grilse ceased eating altogether prior to spawning, which helped cause their meat to be undesirable. After the grilse spawned, almost all died.

In salmon aquaculture, companies tried to replicate the natural cycle of Atlantic salmon with techniques that shortened the total cycle to less than three years. The freshwater phase was replicated in hatcheries with artificial spawning in buckets from genetically selected spawners (adults). Changes in water temperature and amount of light helped the salmon progress quickly through the freshwater stages in eight to twelve months. During this time, water quality and

1 Zones included several sites at which salmon were raised in cages; 10 to 50 cages typically comprised a single farming site, and several sites occupied a single region or zone.

( -2 - ) ( UVA-QA-0749 )

vaccination were the most important factors for future growth and survival. Near the end of the freshwater stage, smolts were moved into estuary sites to improve their ability to adapt to the

( -3- ) ( UVA-QA-0749 )

salinity

of their future seawater homes. In

the wild,

it was estimated that

only 15% of fry

survived the freshwater stage (less than 1% from egg to smolt)—compared with a 90% survival rate in captivity.

The improvement in survival was

even more

pronounced

during the

seawater

stage.

Whereas only about 35% of wild Atlantic salmon survived four years in the open ocean to reach maturity, careful care and feeding in commercial salmon farms led to commercial size in about 18 to 24 months, with 85% survival rates.

The SP Jacaf Fjord farming site was typical of the approximately 800 sites that were

operatin in Chile. It maintained 24 rectangular cages grouped into two rectangular sets of 12

cages containing 900,000 salmon, a floating house for workers, and a floating/movable platform carrying feed silos and blowing machines. Each cage was a floating wharf with walkways along the net boundaries from which a cube-shaped net was suspended and anchored to the sea floor (see Figure 1). At the bottom of the net, a specially designed cone allowed dead fish to be easily removed and counted upon sinking to the bottom of the net.

Figure 1. Typical salmon cage.

Source: Created by case writer.

A crew tended the site around the clock; the crew’s main duties were to feed and protect the fish. Fish were fed a mixture of fish oil, fish meal, and other components twice a day and were allowed to eat as much as they wanted. When conditions were right, the fish ate more and grew faster.

The harvest was one of the most critical operations in the process because it dramatically affected the quality of the finished product. Harvesting at the Jacaf site was conducted by Ace Services using boats with special pumps and a percussive-stun harvest system that killed the fish

instantly with a blow to the he d from a pneumatic piston. The harvest crew (Ace employees

with assistance from SP crews) then bled the fish by cutting the gill arches and placed them in ice water for transport to the SP processing plant in Puerto Aisén, Chile. This single processing

plant served all five of SP’s sites across three zones. (SP had only one site in the Jacaf Fjord zone.)

Harvested salmon were used in the production of three different products in fresh or frozen formats. The production mix was determined based on the weight distribution of the harvested fish. Smaller salmon were sold entero (whole), midsize salmon were sold as “value- added” porciones (portions), and the largest were sold as filetes (filets). The fresh versus frozen mix was decided based on prevailing prices and market demand. SP’s policy (simplified for the purposes of this case) was to freeze 70% of the entero and filetes production and freeze 100% of porciones.

Upon the salmons’ arrival at the processing plant, an automatic grader gutted and sorted them by weight. (See Exhibit 1 for a diagram of the processing plant’s operations and Exhibit 2 for photos.) At the next station, workers removed heads and tails (depending on the product), after which an automatic fileting machine split each fish in two, and another machine removed spines, bones, and skin. At a subsequent workstation, any remaining bones were removed manually. Finally, highly skilled workers trimmed the filetes according to customer requirements. Filetes destined for portioning were put on a portioning machine that scanned and optimized the cuts, converting each filete into four or five porciones.

Fresh products were then packed and shipped the same day. Frozen products passed through a tunnel freezer prior to packaging and were stored for later shipment. Each box of the finished product was printed with codes that identified every aspect of the production of the product in that box.

Salmones Puyuhuapi

SP was a relatively small company, with annual sales of $40 million entirely from salmon sold in one of the three product forms mentioned. (The company did its accounting in U.S. dollars because most sales were to the United States.)

Table 1 lists the current prices and processing costs for each of the three products. A complicating feature of aquaculture economics was the distinction between live and finished product weight. The ratio of the weight of the finished product (after processing) to the weight of the live fish before processing was called the yield. Each product had a separate yield factor that was remarkably consistent across fish and companies. This meant that those in the industry were very comfortable talking in terms of either live or finished weights—and easily interchanging between the two. For example, salmon between 3.6 and 4.0 kilograms were processed as porciones and received (on average) $3.71 per kilogram live weight, or (given the 45% yield)

$8.25 per finished kilogram.

( -4- ) ( U VA-QA-0749 )

Table 1. Production economics.

Product	Live Weight Policy	Yield	Target Weight (kg)	Price at Target (per kg)	Price Change Rate	Processing Cost (per finished kg)
enteros, fresh	< 3.6 kg	0.89	3.20	$3.25	$2.00	$0.45
enteros, frozen	< 3.6 kg	0.89	3.20	$3.20	$2.00	$0.35
porciones, frozen	3.6 to 4 kg	0.45	1.71	$8.25	$1.00	$1.70
filetes, fresh	> 4 kg	0.61	2.75	$5.45	$2.00	$0.70
filetes, frozen	> 4 kg	0.61	2.75	$5.40	$2.00	$0.60

Fish weighing less than 3.6 kilograms at harvest were sold as enteros. The finished weight of a fish sold whole was 89% of its live weight. Fish weighing between 3.6 and 4.0 kilograms at harvest were sold as porciones that, although facing a 45% yield, brought the highest price per finished kilogram of $8.25. (Note that the accounting here is done per fish. A fish produced two filetes, or eight to ten porciones.) Finally, the largest fish at harvest were processed as filetes with finished weights 61% of the live weight.

Of course, not all finished fish weighed exactly the same. The prices listed in Table 1 were for finished products averaging the target weight. Departures (up or down) from that target average weight received prices adjusted at the rate of $2.00 per kilogram (for enteros and filetes) and $1.00 per kilogram (for porciones). Thus if a shipment of fresh whole salmon averaged 3.1 kilograms per fish (0.1 kilograms under the target), SP would receive $3.05 per kilogram. (The actual pricing of salmon products was more complicated because finished product was sold in predefined-size “buckets” with a specified price per kilogram of product meeting the weight standards of each bucket. Richards used the simplified pricing model in Table 1 for projecting revenues.)

The Jacaf Fjord farming site had monthly fixed costs of $50,000 ($20,000 of which were for the crews), and the salmon at this point were eating 1.3 kilograms of feed for each kilogram of weight gained. Feed prices were extremely volatile, with a price of $1,200 per ton (1,000 kilograms). Harvest costs (Ace Services) were approximately $0.2 per live kilogram.

The Meeting

Correa closed the door to his office so that he and Richards and Rivas could speak privately about how to respond to the potential crisis. With his laptop open, Richards did not wait for permission to speak. (Exhibit 4 contains the English translations of their conversation.)

Jorge Richards (operations manager)

Creo que no debiésemos perder tiempo y cosechar la totalidad del centro ahora. Tenemos 900,000 salmones en Jacaf con un peso promedio de 2.77 kilos. ¡Recién hable con el gerente general de Ace Services y estamos con suerte! Un servicio

( -5- ) ( UVA-QA-0749 )

que tenían contratado con otra empresa fue cancelado a último minuto, por lo que tienen tres barcazas con su personal disponible para nosotros. Somos afortunados considerando que ellos generalmente no tienen esta flexibilidad y los servicios se contratan con al menos seis meses de anticipación. Me da terror correr el riesgo de que nos agarre el ISA. El retiro y apropiada disposición de la mortalidad del ISA cuesta casi lo mismo que cosechar y procesar. Los peces están a un peso suficiente para cosechar, procesar y exportar. Puse todos los números en un Excel (Exhibit 3) con mis proyecciones en la actual situación si cosecháramos. Usando una desviación estándar de 0.4 kilos podríamos obtener ingresos por casi $4 millones. Nos cuesta $500 mil cosechar y aproximadamente $1 millón procesar. La contribución seria $2.5 millones—$2,567,376 para ser precisos. Tomemos el servicio de Ace Services ahora, antes que alguien más lo haga o sea demasiado tarde. Por supuesto mi plan solo funciona si logramos convencer a los operarios de la planta trabajar un par de semanas más antes de cerrar la planta por fin de temporada. Estoy seguro de poder hacerlo.

Osvaldo Correa (CEO)

Tranquilo Jorge. No nos apresuremos demasiado. Justamente Sergio me sugería que, por estar en el lado opuesto del centro infectado en el fiordo, hay una probabilidad razonable que podamos salvarnos del virus. Por supuesto estoy considerando las medidas extraordinarias tomadas desde el 2007. Sergio reviso las corrientes, temperaturas y las condiciones del fiordo. El riesgo está latente pero no es 100% seguro que nos infectemos. Incluso, siendo atacados por el virus no necesariamente perderemos todos los salmones del centro. Claramente en el peor escenario perderíamos todos, pero también existe un escenario en que al virus solo mataría al 30% de los salmones. También podemos considerar un escenario intermedio donde solo perdemos al 60%. Y esto aplica si y solo si el centro es infectado por el virus y espero que eso no pase.

Recuerda que en los últimos dos meses previos a la cosecha es cuando el salmón crece más rápido (25% mensual). Y ese 25% en peso promedio y desviación estándar se traduce en un aumento aun mayor en ingresos y utilidades. Por lo mismo, porque no aprovechamos tu planilla y vemos cuanto es lo que estamos dejando en la mesa si decidimos cosechar ahora.

Finalmente veo que tus $2.5 millones es contribución. ¿Cómo puedes ignorar los

$4.2 millones que ya hemos gastado en el centro en términos de 16 meses de costos fijos, costo de smolts y alimento a la fecha? ¿Tus $2.5 millones son en realidad una pérdida de $1.7 millones, cierto?

( -6- ) ( UVA-QA-0749 )

Sergio Rivas (veterinarian)

En el escenario donde nos salvamos del virus, Jorge, por favor considera el 2% de mortalidad habitual de los últimos dos meses. Además recuerda que si el virus nos ataca, los salmones sobrevivientes verán afectados su crecimiento desde un 25% a 20% mensual. La mortalidad del virus tiene que ser retirado por un servicio especial que nos cobra $1 por cada salmón.

La otra opción que podemos ver es la de vacunar los 900,000 salmones inmediatamente. Sé que es bastante costosa la vacuna pero lo podemos hacer de inmediato. Ayer mismo hablaba con el representante de la empresa y tiene vacunas en stock. Incluso Ace Services nos podría ayudar con sus barcazas, plataformas y tripulantes. Con ayuda de una red movemos los peces a un lado de la jaula, los bombeamos, vacunamos y devolvemos al otro lado de la red en la misma jaula. Una vez vacunados no habrá posibilidad de que sean infectados con el virus. Ace Services podrá volver en dos meses y realizar la cosecha previamente programada de salmones con un peso de 4.33 kilos. Jorge, por favor incorpora eso en tu planilla y evalúalo.

The meeting ended with Rivas’s agreeing to collect more information about the outbreak and the likelihood it would spread to SP’s site. Richards agreed to contact Ace Services to get an idea of the feasibility and cost of Rivas’s vaccination plan. Correa volunteered to alert the processing plant there might be another couple of weeks’ worth of work before shutting down for the off-season. He also set about the task of running some numbers to get some understanding of the tradeoffs involved. He was well aware that the ISA virus harmed fish but not humans. Fish that died or were sick were disposed of properly, but fish harvested live (even from an infected cage) could be processed, inspected, and sold without a problem. He also realized that whatever action he took with the current 900,000 fish would have little impact on when SP would be able to start a new cycle at Jacaf Fjord. That would depend mostly on the severity of the outbreak in the zone as a whole. Early harvest or vaccination of his fish would mean very little.

( -7- ) ( UVA-QA-0749 )

-8-

UVA-QA-0749

Exhibit 1

SALMONES PUYUHUAPI (A)

Processing Plant Diagram

Source: Created by case writer.

64

For use only in the course Analytical Decision Making Tools at University of Massachusetts-Lowell taught by Thomas Sloan from January 22, 2019 to May 14, 2019.

Use outside these parameters is a copyright violation.

-9-

Exhibit 2

SALMONES PUYUHUAPI (A)

Selected Photos of Salmon Production1

H

ta

th

UVA-QA-0749

intained in st size of

atchery: Atlantic salmon lived inside these nks until they reached 60 grams; they were en moved to estuary sites.

Sea site: Atlantic salmon were ma cages until they reached a harve approximately 4 kilograms.

Processing plant: Skilled workers removed the remaining bones from filets.

1 All photos were taken by Miguel Angel Leiva and are used with permission.

65

xpo

rted

ant

factor

F

roz

-10-

UVA-QA-0749

Exhibit 2 (continued)

Frozen entero salmon, ready to be e

en filetes

Fresh filetes: Color was an import in the quality of the product.

66

For use only in the course Analytical Decision Making Tools at University of Massachusetts-Lowell taught by Thomas Sloan from January 22, 2019 to May 14, 2019.

Use outside these parameters is a copyright violation.

Exhibit 3

SALMONES PUYUHUAPI (A)

Immediate Harvest Financial Projections (weight in kilograms)

INPUT ASSUMPTIONS
Number of fish	900,000		Harvest cost per live kg	$0.20
Average live weight	2.77		Feed cost per ton	$1,200
Standard deviation	0.4		Feed efficiency	1.3

Product	Yield	Target Weight	Price at Target	Price Change Rate	Process Co (per finished	st kg)
enteros, fresh	0.89	3.20	$3.25	$2.00	$0.45
enteros, frozen	0.89	3.20	$3.20	$2.00	$0.35
porciones, frozen	0.45	1.71	$8.25	$1.00	$1.70
filetes, fresh	0.61	2.75	$5.45	$2.00	$0.70
filetes, frozen	0.61	2.75	$5.40	$2.00	$0.60

( -11- ) ( UVA-QA-0749 )

( 67 )

30%

30%

MATERIAL FLOW

Live Weight	Number of Fish	Average Live Weight	Total Live Weight	Average Finished Weight	Total Finished Weight
< 3.6 - enteros	882,906	2.75	2,428,967	2.45	2,161,781
3.6 to 4 - porciones	16,147	3.72	60,138	1.68	27,062
> 4 - filetes	947	4.11	3,894	2.51	2,376

Total 900,000 2,493,000 2,191,219

PRODUCT FLOW

Product	Number	Average Finished Weight	Price	Revenue	Process Cost
enteros, fresh	264,872	2.45	$1.75	$1,132,968	$291,840
enteros, frozen	618,034	2.45	$1.70	$2,567,930	$529,636
porciones, frozen	16,147	1.68	$8.22	$222,344	$46,006
filetes, fresh	284	2.51	$4.97	$3,539	$499
filetes, frozen	663	2.51	$4.92	$8,174	$998

Total 900,000 $3,934,955 $868,979

FINANCIALS

$ 2,567,376

Contribution

$ 3,934,955

$ -

$ 498,600

$ 868,979

Total revenue Feed cost Harvest cost

Processing cost

Source: Created by case writer.

Exhibit 4

SALMONES PUYUHUAPI (A)

English Translation of the Meeting1

Jorge Richards (Operations Manager)

I don’t think we should waste time and harvest the entire center now. We have 900,000 salmon in Jacaf that weigh 2.77 kilos on average. I recently spoke with the general manager at Ace Services, and we’re in luck! An order that had already been contracted with another firm was canceled at the last minute, and as a result, they have three barges and their personnel available to harvest us. We’re lucky considering that they don’t usually have this much flexibility and these services are booked at least six months in advance. It terrifies me to run the risk that ISA is going to catch us. Proper disposal of the refuse due to ISA costs almost as much as harvesting and processing. The fish weigh enough now to harvest, process, and export. I created an Excel file with my projections if we harvest now under this situation. Using a standard deviation of 0.4 kilograms, we can generate almost

$4 million in revenue. It costs us $500,000 to harvest and approximately $1 million to process. The contribution would be $2.5 million—$2,567,376 to be precise. We should grab Ace Services now before someone else does it or it becomes too late. For my plan to work, I’ll have to convince our plant workers to stay for a couple more weeks of emergency work before shutting down for the off-season. But I can do that.

Osvaldo Correa (CEO)

Take it easy, Jorge. Let’s not rush ourselves too much. Sergio tells me there is a reasonable probability that we can avoid the virus because the infected area is on the opposite side of the fjord. It goes without saying that I am considering the extraordinary measures taken since 2007. Sergio checked the currents, temperatures, and conditions in the fjord. There is some latent risk, but the chance of being infected is not 100%. In addition, being attacked by the virus does not necessarily mean we would lose all the salmon in the center. Clearly, in the worst- case scenario, we would lose them all. But there is also a scenario in which the virus would only kill 30% of the salmon. We can also consider an intermediate scenario in which we would only lose 60%. And this applies if and only if the virus infects the center, which I’m hoping doesn’t happen.

1 Translation was by Raul O. Chao, Assistant Professor, Darden Graduate School of Business Administration.

( -12- ) ( UVA-QA-0749 )

Exhibit 4 (continued)

Recall that the salmon grow fastest (25% monthly) during the last two months prior to harvest. And that 25% in average weight and standard deviation translate into an even bigger gain in revenue and profit. Nonetheless, why don’t we take advantage of your analysis and see how much we’re leaving on the table if we decide to harvest now?

Finally, I see that your $2.5 million is contribution. How can you ignore the

$4.2 million that we’ve already invested in the center in terms of 16 months of fixed costs, cost of smolts, and food to date? Your $2.5 million is actually a loss of $1.7 million, right?

Sergio Rivas (Veterinarian)

Under the scenario in which we avoid the virus, Jorge, you have to consider the usual 2% mortality rate of the last two months. In addition, recall that if the virus attacks us, the salmon that survive will see their growth affected by 25% to 20% monthly. Mortality due to the virus has to be removed by a special service that charges us $1 per salmon.

The other option that we have is to immediately vaccinate the 900,000 salmon. I realize that the vaccine costs a lot, but we can do it right away. Just yesterday I spoke with the representative from the company, and he has vaccine in stock. Also, Ace Services can help us with their barges, platforms, and crew. Using a net, we’ll move the fish to one side of the pen; we’ll pump them, vaccinate them, and we’ll return them to the other side of the net in the same pen. Once vaccinated, there would be no chance that they become infected with the virus. Ace Services can return in two months and have the previously planned salmon harvest with a weight of 4.33 kilograms. Jorge, please include that in your analysis and evaluate it forthwith.

( -13- ) ( UVA-QA-0749 )