Table of Contents

1. Introduction

1.1. Ongoing work in Pisco - general information only

1.2. Project funded by the usaid peru office

1.3. Peru landings and estimated discharge since 1956

1.4. The major key to pollution prevention is raw material quality

2. The process

2.1. Key areas to consider

2.1.1. Raw material quality

2.1.2. Unloading system

2.1.3. Blood water

2.1.4. Stickwater evaporation

3. Critical assumptions regarding the peruvian fishmeal industry

4. Raw material quality

4.1. Discussion

4.1.1. Yield losses

4.1.2. TVN - Biogenic amines

4.1.3. Preservation

5. The unloading process

5.1. Options

5.2. Unloading operation diagram

5.3. Wet unloading

5.4. Dry unloading

6. Blood water

6.1. Definition

6.2. How is it formed

6.3. Processing blood water

7. Stickwater

7.1. Definition

7.2. Composition

7.3. Assumptions on volume

8. Economic considerations

8.1. Raw material quality

8.2. Pump water recovery

8.3. Blood water recovery

8.4. Stickwater recovery

9. Summary of economic losses

10. Course of action

10.1. Objective 1

10.2. Objective 2

10.3. Objective 3

10.4. Objective 4

10.5. Estimated investment costs and returns on capital

11. Other processing methods

11.1. Fish silage

11.2. Composting

11.3. Extruded feeds

11.4. Bio-gas production

11.5. Discharge into the sea

12. Summary

13. Open discussion

List of Figures

Figure 1. Peruvian landings Vs fishmeal landings

Figure 2. Estimated losse in the peruvian fishmeal industry 1950 - 1994

Figure 3. Major key to pollution prevention

Figure 4.

Figure 5.

Figure 6. Key areas to consider in the fishmeal process

Figure 7. When the raw material spoils

Figure 8. Reducing the storage temperature 5ºC - 6ºC

Figure 9.

Figure 10. 5 year average (1991 - 1995) landings of fish at peruvian ports

Figure 11. Critical assumptions

Figure 12. Metric conversion of MG/L dry solids to KG/TON of fish processed

Figure 13. Calculations to convert MG/L in the water to tons of fishmeal

Figure 14. Material mass balance for a typical anchovy fishmeal plant

Figure 15. Herring yield loss when stored at different temperatures

Figure 16. Weight losses in various fish stored at 3 different temperatures over 17 days

Figure 17. Daily losses in weight during bulk storage of fish

Figure 18. Nitrogen X 6.25 = Protein

Figure 19. Increase in TVN of stored capelin over time at different temperatures

Figure 20. Change in TVN content of harring and small cod over 13 day period

Figure 21. Preserve the quality of the raw material

Figure 22. The cumulative effect of raw material quality on the fishmeal process

Figure 23. Criteria for ideal unloading systems

Figure 24. Raw material unloading systems

Figure 25.

Figure 26.

Figure 27.

Figure 28.

Figure 29.

Figure 30.

Figure 31.

Figure 32. Unloading and handling systems for fish

Figure 33.

Figure 34.

Figure 35. Pumpwater flow diagram

Figure 36. Wet unloading process disadvantages

Figure 37. Pumpwater losses calculated as fishmeal by port

Figure 38. Dry unloading process disadvanteges

Figure 39. Blood water losses calculated as fishmeal by port

Figure 40. Blood water process flow diagram

Figure 41. Stickwater process flow diagram

Figure 42. Loss of solids in the pumpwater with and without screening

Figure 43. Loss of solids and their value from not processing the blood water

Figure 44. Loss of solids in the stickwater if only 25% of the factories discharge the stickwater

Figure 45. Summary of yield and economic losses by processing step

Figure 46. Pumpwater challenges

Figure 47. Pumpwater suggestions 1

Figure 48. Recycling of pumpwater

Figure 49. Suggested fish unloading recycle system

Figure 50. Screens

Figure 51. Screen testing results

Figure 52. Blood water recovery

Figure 53. Blood water data generated - Pisco plants

Figure 54. Possible technology combinations and otoions

Figure 55. Investment costs and returns by options

Figure 56. Flow diagram of the silage production process

Figure 57. Simpe fish waste composting system

Figure 58. A sick compost pile 1

Figure 59. A sick compost pile 2

Figure 60. A sick compost pile 3

Figure 61. A sick compost pile 4

Figure 62. The environmental dilemma

1. Introduction

1.1. Ongoing work in Pisco - general information only

1.2. Project funded by the usaid peru office

1.3. Perú landings and estimated discharge since 1956

1.4. The major key to pollution prevention is raw material quality

The title of this presentation suggests that seafood processing will be the main topic but as you can see from Figure 1, fishmeal is the dominant industry in the seafood area in Perú and will be the subject of this presentation. Pollution prevention in other segments of the seafood industry parallel those in the fishmeal industry. In fact, if you process seafood for edible use you almost always will produce solid wastes which must be either converted to fishmeal or handled in some other manner. Waste waters generated in the edible sector of the industry, although probably in lesser volumes will still have many of the same characteristics as those generated in fishmeal plants. Therefore, the technology can be transferred. We will cover some aspects of handling fish and shellfish wastes other than fishmeal. These other processes are inexpensive to set up but might require a heavy commitment to market development.

In order to give you some idea of the magnitude of the problem and the potential for increased revenues, we have estimated that between 1950 and 1994 the Peruvian fishmeal industry has landed 199 million metric tons (mmt) of fish (anchovy and sardine). At a conversion of 6:1 fish to fishmeal, that converts into 33.3 mmt of fishmeal. If we use an 11 year average CIF Hamburg price US$431 for fishmeal less US$60/ton for freight, we arrive at a figure of US$371/mt or US$12.35 billion (thousand million) for the value of the fishmeal produced over that period of time. Based on the average losses that we have found in Pisco in only the pumpwater, we have estimated that potentially about 6.85 mmt of fishmeal were lost over that period. At the average fishmeal price this works out to US$2.54 billion (thousand million) or about 20.6% of the fishmeal produced (Figure 2). We understand that the fishmeal industry has evolved over this time period, we realize that average numbers may not be the best way to evaluate the situation and that every factory is different. We are saying that even if we reduce the numbers by a factor of 100 in our calculations, the figure is still very large and presents some unique opportunities to the industry for enhancing yields. We will be going into the details on how we arrived at these figures and what steps are needed to evaluate each facility during today's presentation.

The major key to pollution prevention in the fishmeal industry is the quality of the raw material (Figure 3). Raw fish quality affects every stage of the process from storage on the vessel through evaporation of the stickwater. As soon as the fish are captured they begin to deteriorate and you start to lose yield on the vessel. By the time the catch is processed, you have not only lost valuable product but the quality of what you have produced is reduced. Yield loss and lower quality equate to reduced value for the products produced and higher production costs. Lower yields and reduced quality make you less competitive in a market that is moving towards prime and super prime fishmeals.

2. The process

2.1. Key areas to consider

Before we can assess the sources of product loss and therefore pollution, we must understand the process. The basic process for producing fishmeal is outlined in the attached flow diagram (Figure 4). This is a generic flow diagram and I am sure it will vary for each plant in Peru. When we made our assessment of the Pisco plants, we started with this diagram and a two-page questionnaire. Based on the answers in the questionnaire, we modified the flow diagram for each plant. The interview took about two hours at each plant. The questionnaire that we used is shown in Figure 5. The diagram and questionnaire are key elements of a pollution prevention program and are necessary for doing an assessment of the facility.

2.1.1. Raw material quality

2.1.2. Unloading system

2.1.3. Blood water

2.1.4. Stickwater evaporation Figure 6.

1. Reduces the quantity of products (yield)
2. Increases the cost of production
3. Causes pollution problems
4. Reduces quality of the finished products. Figure 7.

You cannot make a prime or high quality product if the raw material is not fresh. You cannot expect to recover the optimum yield from the raw material if solids and oil are going out in the vessel bilge, in the pumpwater water, in the blood water and in the stickwater if it is not evaporated. If the fish are stale, the stickwater will be stale and the concentrate added back on the presscake will reduce the quality of the fishmeal further. The volatiles from the stickwater will come out and enter the condensate water raising the BOD}5} and when the presscake is dried the volatiles in the fish will be discharged to the atmosphere causing odors. Therefore your first line of defense is fresh raw material. Reducing the storage temperature of the fish by only 5-6 centigrades will reduce the biochemical reactions that cause spoilage by 50% and extend the storage time by 100%. Figure 8.

Bloodwater is produced on the fishing vessel when the catch is stored during the return trip to the factory and after the fish have been unloaded into the raw bins of the plant. Blood water results from bacterial activity and autolysis (self digestion) from the enzymes in the fish gut and from what the fish are eating. This reaction increases with temperature and both protein and oil are lost as a result. Once the fish reach the raw holding bins they continue to deteriorate. Liquids (oil and water with solids) continue to leach out of the fish. If not processed, valuable product is lost. A report from Chile states that the solids in the bloodwater increased from 5% after 1 hour storage to 14.5% after 21.5 hours of storage. A Scandinavian report indicated that the bloodwater losses could amount to 10-15% of the original weight of the raw material. Figure 9. The bloodwater is produced by the decomposition of the fish and is released by the pressure that they undergo in storage. If the produced bloodwater is not allowed to escape from the fish, it will accelerate the decomposition process and produce more bloodwater.

3. Critical assumptions regarding the peruvian fishmeal industry

In order for us to put meaningful values on the unit operations that result in yield loss, it is necessary to make some assumptions (guesses). These are educated guesses that have been further modified by using very conservative figures. You might not agree with these so we should discuss them a little later in the presentation. These assumptions are outlined in Figure 10 and Figure 11.

3.1 % OF PRODUCTION BY CITY 3.2 5-11 YEAR AVERAGE FAQ FISHMEAL PRICE, HAMBURG LESS US$60 3.3 5-44 YEAR AVERAGE PERUVIAN LANDINGS 3.4 5-11 YEAR AVERAGE CRUDE FISH OIL PRICE, ROTTERDAM LESS US$60 3.5 HIDROSTAL PUMP 2:1 RATIO OF WATER TO FISH 3.6 PROTEIN + OIL IN DISCHARGE = FISH SOLIDS 3.7 CALCULATIONS

Several formula were developed in the EP3 project to convert data into tons of fish, tons of fishmeal and the value of the fishmeal. These are outlined in Figure 12 and Figure 13.

3.7.1 METRIC CONVERSION MG/L TO KG/TON 3.7.2 MG/L SOLIDS/500 = KG DRY SOLIDS LOST PER MT FISH 2:1 RATIO 3.7.3 MG/L SOLIDS/1000 = KG DRY SOLIDS LOST PER MT FISH 1:1 RATIO 3.7.4 5:1 CONVERSION OF RAW FISH TO FISH MEAL 3.7.5 KG DRY SOLIDS X 1.11 = KG FISH MEAL AT 10% MOISTURE 3.7.6 KG OF FISHMEAL LOST/TON OF FISH X $381/TON = REVENUE LOST US$ 3.7.7 ASSUME 2000 HOURS PER SEASON OF OPERATION AT A CAPACITY OF 50 TONS OF FISH/HOUR (1 LINE)

If we assume 2000 hours per season of operation at a capacity of 50 tons/hour, we come up with a plant that processes 100,000 metric tons of fish per season. Figure 14 gives the expected material balance for a typical plant. It is based on 1000 kg of fish and assumes that there are no losses.

4. Raw material quality

4.1. Discussion

We can see from the above calculations that raw material quality is very important. Fresh fish lose much less solids to the pumpwater than stale fish. In fact, for every 100 mg/l of protein + fat in the pumpwater, you will lose 0.22 kg of fishmeal per ton of fish.

4.1.1. Yield losses

4.1.2. TVN - Biogenic amines

4.1.3. Preservation

Based on research done in the fishmeal industry in other countries, we found that the pressure of the fish on each layer in the hold is one of the primary causes of raw material deterioration. The more fish in the hold, the more quickly the pressure will force the fish to compress and lose liquid. This liquid contains enzymes which will further digest (liquify) the mass producing more liquid. Deterioration increases with increased storage temperatures and time. While the data from other species such as sandeel, herring, pilchards and sardines may not be directly related to anchovy, the principle is the same; when the fish are stored at elevated temperatures under pressure, they will lose liquid which if not recovered leads to reduced yields and pollution of the surrounding waters.

For herring it was calculated that for each 5 centigrades increase in temperature, the daily loss of yield doubles. Figure 15. Work done with pilchards showed an increase in solids in the blood water from 1.6% to 11.4% in 2 days at 15 centigrades. With broken fish or trimmings, the bloodwater weight increased from 2.6% in 6 hours to 5.06% in 24 hours and the lost solids went from 1% in 6 hours to 1.75% in 24 hours. The blood water can range from 10-15% of the weight of the raw material. Studies with cod, redfish and herring compared the weight losses over 17 days at three different temperatures. Figure 16. Danish studies with sandeels indicate that the loss in solids in the vessel bilge water discharged at sea could amount to as much as 3.5% of the weight of the fish. In Norway studies with herring and Norway Pout stored at three different temperatures, showed similar losses in weight when the temperature was increased. Figure 17.

Spoilage of fish can be measured in a number of ways. The easiest way is to use your nose. The spoiled odor is due to compounds that form when bacteria and chemicals in the fish begin to breakdown the proteins. They can also increase the FFA in the oil. The resulting compounds are called volatile nitrogen compounds and biogenic amines. Ammonia is a volatile nitrogen compound and histamine is a biogenic amine. The volatile nitrogen compounds are usually measured as TVN. The analysis of the biogenic amines is more complicated and requires expensive instrumentation. When these compounds form, they come out of the fish and are lost in the stickwater, blood water and evaporator, and dryer condensates. For every 100 TVN units you lose 0.625% protein in the fish. Figure 18. Fresh fish should have a TVN of 10-15 mg/100 g of fish.

Norwegian studies with capelin showed large increases in TVN and oil FFA when the fish were stored in bulk. Figure 19. Danish studies with herring and mackerel showed that the biogenic amine content (histamine, cadaverine putrescine and spermidine) were 2-20 times higher in the fish at 10 centigrades than at 2 centigrades. In another Danish study the TVN content of small cod and herring was measured at three different temperatures over two-weeks storage. The lower temperatures resulted in increased storage times. Figure 20.

If the catch could be processed on-board the vessel immediately after capture, there would be no need for preservation techniques but this is not the case in the fishmeal industry. There have been many studies done to evaluate methods of preserving fish for edible use and fishmeal production. These include protection from contamination, icing, containerization, refrigerated seawater, and chemicals. Figure 21. To reduce spoilage of the fish you must eliminate or reduce bacterial contamination and lower the temperature of the fish. You must also prevent the fish from crushing each other so that the enzymes in the fish do not come out and begin digesting the proteins.

Anchovy present problems when it comes to preservation. They are small, the belly's burst very quickly after capture and they compress in the hold of the vessel. Experiments have been conducted with the "champaign" system of refrigeration where the chilled water is forced up through the fish rather than sprayed from the top. It appears that major modifications of the fleet will be necessary to effect some type of preservation of the fish. This might not be possible as retrofitting the existing fleet might be too costly or impossible. Such modifications should include insulation of the holds, compartmenting the holds so that fewer fish will be in each compartment and possibly the use of containers to further reduce the volume in each compartment.

4.1.3.1. Advantages of preservation

The advantages to preservation of the catch will be increased yields and higher quality fishmeal and oil. Fresh raw material carries through the process like money in the bank increasing with time. Figure 22. You can't collect on what you don't have.

4.1.3.2. Disadvantages

There are disadvantages to preservation of the raw material. Nothing is free. There will be increased capital costs in the construction of the vessels.

4.1.3.2.1 Reduced vessel carrying capacity

The vessels will carry less fish because the walls of the hold must be insulated, so the cost to catch a unit of fish will increase.

5. The unloading process

5.1. Options

5.2. Unloading operation diagram

5.3. Wet unloading

5.4. Dry unloading

 

In our assessment of the plants in Pisco we found that the first place in the process where effluent and therefore product was being lost was in the unloading process. We contacted fishmeal producers in most of the major countries to determine what unloading methods were being used and what might be applicable to the Peruvian industry.

The discharge of industrial fish from the vessel to the processing facility has presented many challenges to the fishing industry over many years. The methods employed must not only be economical and pollution-free, but they must get the fish to the plant in good shape as quickly as possible, since the vessel must be able to get back on the fishing grounds while fish are available. Global environmental regulations in the late 1960's and early 1970's accelerated the development of new methods for discharging fish to the factory so that the various companies could meet the new developing environmental standards. In 1977 the International Association of Fish Meal Manufacturers (IAFMM now IFOMA) organized a symposium to cover this issue. Speakers were invited to address the problem of getting the fish from the vessel to the factory. The keynote paper outlined the Norwegian requirements for an unloading system that would address strict anti-pollution requirements, restricted working hours, and high labor and energy costs. These requirements are still valid today and are listed in Figure 23.

Generally speaking, there are only two ways to get the fish from the fishing vessel to the factory for processing. The fish can be moved either wet or dry. Within these two categories, there are at least seven options for unloading fish from the vessel to the plant that are or have been used in the fishing industry. These are outlined in Figure 24.

5.1. Options

5.2. Unloading operation diagram

5.3. Wet unloading

5.4. Dry unloading

5.1.1.1 Hidrostal PUMP 2:1

The Hidrostal pump has a special centrifugal screw impeller that was developed in Peru specifically for pumping fish. It is the wet pump that is now used throughout the Peruvian fishery and in some plants in Chile. The only connections between the vessel and the shore or platform is a flexible suction pipe and water hoses so there are no problems with tidal movements. The standard pump has a capacity of 50-100 tons of fish/hour and can be varied by adjusting the fish to water ratio.

5.1.1.2 Netzsch PUMP 0.1:1

The Nemo or Netzsch pump is a Mono pump which has been used in various factories for pumping liquids and semisolid materials. The pump consists of a metallic rotor and elastic stator. It is a positive displacement pump with the quantity of material pumped proportional to the speed of the pump. The pump is variable speed and reversible so that it can be cleaned out. The maximum quantity is about 250 cubic metes/hour. For unloading fish, the pump must be moved around the hold of the vessel. This can be done with a crane. This pump is being evaluated at several plants in Peru. One of the disadvantages of the pump is its weight which makes it difficult to maneuver.

5.1.1.3 Pressure/vacuum PUMP 1:1

Recent experiments with a pressure/vacuum pump have been conducted in Chile. The project developed because fish were being discharged from the vessels at the congested port and then trucked from the port through the city to the factories. The pump is reported to be capable of moving fish approximately 1600 meters with less water and less degradation. The pipelines are of high density polyethylene thermofused together and floating on the surface. Fish were conveyed a distance of 1150 meters in the water and 450 meters on land to the plant. The capacity of the pump was 200 tons/hour. See Figure 25. The pumps are electrically powered so power cables must be run to the unloading station. The pump has been successful, unloading 35,000 tons of fish with a ratio of 1:1 to 1:3 of fish to water. Degradation of the fish is no worse than with other unloading methods. These pumps are now operating in Peru and Chile on a variety of species including anchovy.

5.1.1.4 Others

The Superfos Hydraulic transport pump was developed in Denmark in 1973. In looking for an alternative to the wet system, their requirements were:

1. Minimum destruction of raw material.
2. Transportation without the addition of water.
3. Closed system.
4. Minimum maintenance.

The pump is a double-acting piston pump with a four-way valve as the central point. The action of the pistons and the rotary motion of the valve flap allow only one motion to happen at a time. No air or water is used and the actual capacity is 60-80 cubic meters/hour.

In the USA, fishmeal plants use Humphreys piston pumps to unload the fish. These are similar to the Hidrostal pump in that water is used to convey the fish from the vessel into the factory but unlike other systems, the water is screened and recycled in the US factories and finally evaporated to become part of the finished product. By recycling the pumpwater the US plants are able to maintain a 1:3 water to fish ratio.

5.1.2 Dry

The bucket elevator (Figure 26) consists of a bucket that moves the fish to a conveyor, which elevates the fish out of the hold of the vessel and finally to a receiving bin in the factory. While it is less likely to cause pollution of the harbor and surrounding waters, it is affected by tidal changes, cannot be installed on the fishing vessel and is labor intensive. The chainpump-elevator system was used in Denmark. It acts like a pump when the fish are soft and as a bucket elevator when the fish are fresh and firm. It was replaced by the grab because of problems with large tidal changes in the North Sea and North Atlantic.

The Grab (Figure 27) consists of a clamshell or grab at the end of a crane which is attached to a dock. Its disadvantages are the spillage and drippage of fish and liquid onto the dock and into the surrounding waters. It is labor intensive and requires some other method to remove the last traces of fish from the vessel hold. Neither the Grab nor the bucket elevator are used by the fishmeal industry today. Both have been replaced by more environmentally and labor friendly methods.

5.1.2.1 Myrens pump

The Myrens dry pump offers two alternatives to mounting. The pump can be handled by a crane and lowered into the hold of the vessel (Figure 28) or mounted on the vessel (Figure 29 and Figure 30). The pump is a positive displacement pump with a rotating valve. The system operates dry, except for the water that is in the hold of the vessel with the fish. No extra transport water is used. At 45 rpm the pump will move 70-80 cubic meters per hour. Pump bearings and seals are of plastic. The Myrens Pump is currently used in Iceland with excellent results. Only a small amount of water (10%) is needed to get the pump started. Pumps of this type are also used in the factory for moving fish around the plant. The Myrens pump is no longer manufectured, but an Icelandic company has just negotiated the rights to manufacture the pump.

5.1.2.2 Iras system

A pneumatic off loading system (IRAS System) (Figure 31) that was designed to handle different species as well as different degrees of freshness and quality was developed in Denmark. The fish and air are sucked from the hold to the separation section where the fish slide down a tube which is closed by a flap valve. When the weight of the fish in the tube is large enough to overcome the vacuum, the flap opens and the fish slide out. The air escapes from the separating section to a cyclone where smaller particles of fish are collected. The air is then either discharged or recycled back to the fish again. Capacities of 1-2 tons/minute on a 200 HP plant have been achieved.

The unit is maneuverable and the hose can be moved from hatch to hatch. The noise level is also reduced. The IRAS System is used in Denmark today. Figure 32.

5.1.2.3. South African System

A dry off loading system is being used in South Africa today (Figure 33). They have unloaded an average of 50-100 tons anchovy/hour with breakage in the range of 2-3%. The fish enter the system by a suction nozzle and are conveyed through pipes to the separator. The fish are discharged from the separator through a slide box valve. According to the author, power requirements of 0.7 to 2.5 horsepower/ton of fish are needed. Figure 34 shows a slide valve instead of a rotating valve.

5.1.2.4 Containers

Some research projects were done in Chile. Fishing vessels that bring in fish for food use and also fish for fishmeal use can be handled using containers and an icing and recirculating technique developed in the UK. Containers are stacked in the central part of the fish hold and reach from the floor to the deck. The containers are separated from the rest of the hold by walls of wood or aluminum. The containers are prefilled with ice before the vessel departs for the fishing grounds, an equal amount of sea water is added and the air circulation is started before filling up with fish. The containers are then removed from the vessel by a crane.

Based on our assessment of the Pisco plants and brief visits to plants in Chimbote and Coishco the current unloading operation is seen in Figure 35. Screens may or may not be used to recover solids from the water before discharge back to the sea.

As environmental issued continue to take the newspaper headlines and as the more aggressive environmental groups put pressure on the fishing industry through consumers, the discharge of the pump water has become a very critical issue and some companies have begun to look at different methods to deal with the problem. The wet unloading process has some major disadvantages (Figure 36).

5.3.1.1 Volume of water

In our typical plant (50 tons/hour - 2000 hours per season) which uses the Hydrostil pump, approximately 200,000 tons or 52.9 million gallons of pumpwater will be produced in a season. For a larger plant, 100 tons/hour, 400,000 tons or 105.8 million gallons of pumpwater will be produced. If you multiply the single plant volume by the number of fctories in the area, the numbers can get very large. The pumpwater can contain as much as 5.7% total solids if it is not screened or 5.1% total solids if it is screened. At this time we have no data available on the semi-wet pumps to compare with the Hidrostal.

5.3.1.2 Salt content

The total solids content of the water is difficult to deal with because the Pacific Ocean water contains over 3% salt. But the salt is in the pumpwater and must be dealt with or eliminated.

5.3.1.3 Yield losses

We mentioned that for the Pisco project we decided to use the Protein + Oil content of the effluent instead of total solids or salt free solids because the protein and oil can only be coming from the fish (unless it is coming from your neighbor's effluent). If we look at the protein + oil content of the pumpwater we find that it will have an average of 2.6% if the water is not screened (51 kg/ton of fish) or 1.6% (31 kg/ton of fish) if it is screened. This means that just screening through a 1 mm screen will recover about 38.5% of the solids in the water. The 1.6% solids that is discharged in the effluent (15657 mg/l) works out according to our calculation to 34.8 kg of fish meal per ton of fish. With a value of $381/ton for the fish meal, this equates to US$13.24/ton of fish processed that is lost. Figure 37 outlines the tons of fishmeal that could be lost for each of the ports both with and without screening. Data is based on a five-year average for landings and fishmeal price.

We have defined dry unloading as the system that uses no water or just enough water to seal a valve. The grab and crane and bucket elevator system will not be discussed since they are not practical anymore except in special cases. The Icelandic Myrens pump might not be able to move the fish the distances required and would need several booster pumps along the line to keep the fish moving. Several people have mentioned that the dry and semi dry pumps cannot handle anchovy because they dry out and plug the line causing a shutdown of the unloading operation and others have mentioned that some of the semi-dry pumps are very heavy and difficult to manage. Besides these issues, dry unloading has some other disadvantages which are shown in Figure 38.

5.4.1.1. Logistics

The disadvantage of the vacuum unloading system is the limited distance that the fish can be transported. If the vessel cannot tie up adjacent to the factory (such as in many plants in South America) so that the discharge of fish is conveyed into the raw holding bins, then it would be necessary to install the unloading systems in a harbor, unload the fish at the port and deliver them to the factory by truck. This was the case with a factory in Mexico (no longer in operation) and has been tried at several locations in Chile. It is better suited to small plants because with larger plants the volume of fish would require many trucks running from the unloading area back to the factory. The other alternative is to construct piers out to the unloading stations, install the dry unloading system on the pier and run conveyors from the station into the plant.

5.4.1.2. Capital cost

Vacuum unloaders are very expensive and the cost of installation of a pier with conveyors 1500 meters out to the stations would be very expensive.

6. Blood water

6.1. Definition

6.2. How is it formed

6.3. Processing blood water

German studies put the loss at about 3.6% of the raw material, while Danish calculations indicated that the annual loss of weight in the blood water was almost equal to the average monthly catch of fish. Polish researchers reported that the blood water ranges from 10-15% of the original weight of the fish. To improve the yields of meal and maintain quality, it is essential to process the blood water at the same time that the fish are being processed. The best solution is to avoid as far as possible the formation of the blood water, by having suitable storage conditions and adequate processing capacity. The drainage of blood water to waste, in any event must be prevented for reasons of hygiene, since it results in great contamination of streams or harbors. In Figure 39 we have calculated the amount of fishmeal that is being lost if blood water is not recovered. We have done this for each of the ports in Peru, making the assumption that 20 kg/ton of fish is being lost in the bloodwater (equivalent to 2% of the catch, a very low estimation).

Blood water is a mixture of suspended, and dissolved solids plus oil that is released from the fish in storage. There are three options for handling this fraction; discharge overboard, meter back to the cooker for processing with the fish or processing by coagulation of the proteins, separation of the oil and solids and evaporation of the water phase. Figure 40 shows the two alternative treatments for the blood water.

7. Stickwater

7.1. Definition

7.2. Composition

7.3. Assumptions on volume

Fish entering the process are cooked and pressed to separate the solids from the liquids. The solids eventually become fishmeal while the liquids undergo further processing first to recover suspended solids that might have escaped the press and then to separate and recover the oil. The water that is left after the fish oil is recovered is called stickwater.

Waste heat evaporators operate by pumping hot water through a heat exchanger in which the energy is transferred to a circulating flow of stickwater. The stick water is thereby heated and then flashed into a vacuum chamber in which the absorbed energy is flashed off as water vapor. In a single effect plant the vapors are condensed in a sea water cooled condenser. The cooled water is recycled to the condensation tower and again reheated. A single effect wast heat evaporator will generally take about 40-50% of the required water evaporation from the stick water. This reduces the load on the stickwater plant and allows the operator to raise the temperature of the last effect which is necessary if these vapors are to be used in waste cookers. Waste heat evaporators might be sufficient for small operations that cannot afford a multi-stage evaporator. In Figure 41 we have outlined the flow for the liquid phase of the fishmeal process through the production of stickwater concentrate.

As an assumption for this project, about 60% of the fish weight will be generated as stickwater. In Figure 14 we outlined what we considered the theoretical material balance for an anchovy fishmeal plant. We used 200 kg/ton for dry solids and 90 kg/ton for fat. The theoretical yields would be 23.1% fishmeal and 7.6% oil. For our average plant of 50 tons/hour, about 30 tons of stickwater would be produced per hour or 60,000 tons in a 2000 hour season.

8. Economic considerations

8.1. Raw material quality

8.2. Pump water recovery

8.3. Blood water recovery

8.4. Stickwater recovery

We have now reviewed the process. We have isolated four areas of the process for consideration; raw material quality, pump water, blood water and stickwater where it might be possible to recover additional yield. Let's now look at the potential significance of the losses from these different areas.

The next stage in the process where there is a significant yield loss and therefore pollution of the surrounding bays and harbors is discharge of the pump water. With an average 5 year catch of 8.4 million metric tons of fish, the possibility exists that at a ratio of 2:1 water to fish, 16.8 million cubic meters of pump water is discharged into the Peruvian harbors in the average year. Based on the experience at the Pisco plants, and using our assumptions from Figure 11 we can calculate how much product is being lost and what the value of that product might be, based on average data. In Figure 42 we have broken down the Peruvian catch by port and have given the 5 year average catch for each of the ports. We then calculated how much fishmeal is being discharged in the pumpwater both before and after screening at each location. The difference between the two figures is how much additional revenue can be generated if 1 mm screens were installed in each factory. The average 50 ton/hour plant will generate 200,000 cubic meters of pumpwater per year with an estimated 11,322 metric tons of fish meal in it. The 1 mm screen will recover 4,440 metric tons of fishmeal worth $1.7 million in the average year. Screening through a 1 mm screen would be considered primary treatment. There are many types of screens available on the market today. One such device is the rotating hydro sieve which rotates opposite to the liquid flow that is introduced at a tangential angle creating a higher shear velocity. The milliscreen separates small suspended particles from the water and has been successfully used in many applications including the fishmeal industry. These screens can be obtained with openings as small as 0.25 mm.

The next source of yield loss and therefore pollution of the surrounding bays and harbors is discharged of the blood water. Blood water forms when the fish are stored in the raw holding bins. Pressure of the fish releases the liquid which contains oil and proteins. Based on our assumptions and the data published in the literature for other species, we have asumed that 20 kg of solids/ton of fish are lost per day and that the catch is all processed within one day. In Figure 43 we have calculated the losses from bloodwater for each location. While not as substantial as the losses from the pumpwater, the yield loss from discharging the bloodwater is still very significant. Again, for our average 50 ton/hour plant, the losses can be as large as 1110 tons of fishmeal worth an average of US$352,980 in a season.

In the USA and many of the other countries where fishmeal is produced, when the environmental regulations were being developed, the government first looked at what would be considered the best practical technology available to the industry. Best practical technology included the requirement for a stickwater plant or access to a stickwater plant. Fish contain about 80% liquid which consists of water plus oil. As the oil content of the fish increases, the water decreases and vice versa. Based on the material balance that was previously shown, stickwater represents about 60% of the weight of the fish that are unloaded and will contain about 8% solids. For plants that do not currently recover the stickwater, the losses in yield are enormous. Again, using our typical 50 ton/hour plant, about 30 tons of stickwater will be generated per hour. In a season, 60,000 tons of stickwater will be lost and at 8% solids, this works out to 5328 tons of fishmeal worth US$1.69 million. It also represents about 27% of the fishmeal that could be produced from the fish. In Figure 44 we made the assumption that 25% of the Peruvian plants do not have stickwater plants and then calculated the losses in fishmeal value for each of the ports.

9. Summary of economic losses

We've now looked at the potential losses that can result from poor raw material quality, and discharge of the pumpwater, bloodwater and stickwater. Figure 45 summarizes the yield and economic losses from three of these areas and moves us to the next step.

10. Course of action

10.1. Objective 1

10.2. Objective 2

10.3. Objective 3

10.4. Objective 4

10.5. Estimated investment costs andreturns on capital

 

What must we do to recover the losses that result from the discharge of these materials, will the recovery be cost effective and will there be a relatively quick payback of the capital investment. Or are we better off installing a marine outfall pipe and sending everything back to the sea?

10.1.1 Install evaporation plant

From Figure 45 we see that the major source of yield loss and pollution is the loss of the stickwater. Stickwater contributes about 48 kg of dry solids per ton of fish landed or about 19% of the dry matter in the fish. There is a very quick payback when an evaporation plant is installed. There are other considerations, however. The increased dryer load by the addition of the stickwater concentrate could require more dryer capacity. If the raw material is of poor quality, then the volatiles in the water will evaporate and come out in the condensate water and some will carry over into the meal. If the factory does not have steam dryers, the evaporator cannot take advantage of the available waste heat and it would be necessary to add additional boiler capacity. There are opportunities also. In the 1960's, in North Carolina USA, five companies operated fishmeal plants for a relatively short fall season (8-10 weeks). Much of the equipment for these plants was installed for the season by relocating it from factories whose season was over. None of the plants had stickwater plants so the pumpwater, blood water and stickwater were discharged overboard. A new company formed to handle these waters. Evaporators and tanks were installed and an arrangement was made to pickup the water for processing with no cost for the mateiral. A market was developed with companies that took the concentrate and dried it back on vegetable carriers such as alfalfa, soybean meal, and wheat. This new product began to compete with the fishmeal because it was cheaper and eventually the fishmeal plants began to add the solubles back on their own presscake. Most of the factories in that area are now closed but that was the start of wholemeal production in the USA. In Peru, several factories could join together, install and operate an evaporation plant and recover what they are jointly discharging. Or a new company could form to handle this valuable material. If there is no evaporation plant, recovery of the blood water will not be done as it is necessary to evaporate the blood water to recover all the solids so the losses are even higher.

10.2.1 Eliminate or redice water in the unloading process

Pumpwater is the next major source of yield loss and is almost on an equal footing with the stickwater. The pumpwater is further complicated by the fact that: (Figure 46)

1. The source of the water is sea water so there is a very high salt content
2. The volume is very large
3. It is generated in a very short period of time
4. It can contain large amounts of solids and fat.

There may be other issues but these should be the major ones.

We can solve the salt problem by using fresh water if it were available, but it is our understanding that most of the fresh water wells would go dry or there would be salt water infusion into the wells if that volume was removed (Figure 47). For our 50 ton/hour plant we are talking about a minimum of 200,000 cubic meters of water over the season but during the typical unloading operation this could be 400 cubic meters per hour. For all of Peru, the volume of pumpwater should be 16.8 million cubic meters over the season. Another way to eliminate the salt problem would be to go to vacuum unloaders. These unloading systems use air instead of water. A small volume of water is used to seal the valve and for cleanup but other than that there is no water. Vacuum unloaders cannot move the fish the distances that are needed in Peru. In order for the vacuum unloader to work properly it would be necessary to either build a pier out to the unloading station and dry convey the fish to the plant or to install the unloading operation in the City's port so that the vessels can come alongside the dock. It would then be necessary to move the fish to the factory by truck. The dry unloading operation has the advantage that you no longer will worry about yield losses in the pumpwater or long range effluent limitations. But the systems are expensive and the payback might be five or more years.

If we don't have a sufficient volume of fresh water and dry unloading is not practical, we should look at reducing the volume of water. We can reduce the volume of water by going to dry pumps or semi-dry pumps. There are several pumps on the market being used here in Peru and Chile, as well as in other parts of the world. The current pump being used throughout the fishmeal industry here in Peru uses at least a 2:1 ratio of water to fish (this might be as high as 3.5:1).

The pressure/vacuum pump is being used in Chile and several are now being installed in Peru. The pump operates on a ratio of 1:1 and this would reduce the volume of water by a factor of 2. In some cases and with some species of fish the ratio can be reduced. There have been some questions about whether the pump can unload 200 tons/hour and move the fish the required distance (1000-1500 meters). We have no data on the composition of the pumpwater generated by this pump at this time.

The Netzsch pump operates on a ratio of 0.5:1 and this would reduce the volume of water by a factor of 4. This pump is being used in Peru but there have been problems with it. It is very heavy and difficult to move and the capacity is not sufficient for the larger factories. At this point we have no data on the pumpwater from this pump but hope to have it at some time.

The Myrens pump is used in Iceland and Norway. It is a dry pump that operates on a ratio of 0.1:1 and would reduce the water volume by a factor of 25. The Myrens pump is also used to move fish within the factory so that conveyors are not needed. In Iceland they use some of the blood water to start the unloading process. The pump is suspended from a crane and lowered into the hold of the vessel or it can be mounted on the vessel. The pump may not be able to move the fish the 1000-1500 meter distance that is necessary, so it was recommended that three pumps be used in series to move the fish that distance.

If we can reduce the volume of water sufficiently, then we might be able to change over to fresh water or at least process this reduced volume of water.

Recycling of the pumpwater is used in the USA, and some factories in Chile. In the recycle process wet pumps are used at the normal ratio of water to fish. The pumped fish are transported to the factory where the water and fish are separated and the fish are measured or weighed. The pumpwater is then screened through 1 mm screens and collected in a feed tank for recycling back to the vessel. The process is continued until the water is too thick or the vessels have been unloaded.

Recycling (Figure 48) offers several advantages; any system can be retrofitted, the solids buildup is constantly recovered, it uses much less water (in the USA the ratio is about 0.3:1 water to fish) which can then be evaporated. There are some precautions that must be taken in the recycling to prevent the formation of gases, and fresh water must be used to eliminate the salt problem. Depending upon the plant configuration, it might be possible to utilize a waste heat evaporator to concentrate the pumpwater instead of a second evaporator. Figure 49 shows a typical recycling system that might be used in the Peruvian fishmeal industry.

10.3.1 Recover solids from the reduced volume pumpwater

From the literature (Figure 50) and tests in Pisco (Figure 51) we have found that screening the water through 1 mm screens will only recover about 38% of the solids in the pumpwater. Finer screens might be able to recover more, but will not recover the dissolved solids that are in the pumpwater.

The fishmeal industry has experimented with dissolved air floatation (DAF) and dynamic air floatation (DYAF) as a way to recover solids and fat from the waste water streams. DAF involves injecting micro bubbles of air into the screened water so that fine solids and oil are floated to the surface and skimmed off. The water is then discharged and the solids/oil mixture further processed. Using the DAF system instead of a fine screen and then recycling the water back to the vessel, might be another way to reduce water and recover solids. The final water can then be screened or processed along with the stickwater. It should be pointed out that DAF will not remove dissolved solids. Studies have been done over a very long period of time to use coagulating chemicals to recover some of the dissolved solids but then you must deal with the chemicals and the regulatory issues that govern their discharge.

For plants that have operating stickwater plants, the easiest way to recover all the solids from the reduced volume pumpwater would be evaporation. Evaporation is not an option if salt water is used for pumping the fish since the salt content of the water would be further concentrated almost to brine. This brings us back to the fresh water issue. If we can reduce the water volume sufficiently, then it might be possible to utilize fresh water with a recycle system.

10.4.1 Recover the blood water

Bloodwater is the liquid produced during storage of the fish. It is made up of blood from the raw material, some fish solids, plus seawater found in the fish and some pump water. The composition of the bloodwater will vary with the composition of the raw material and the length of time that the fish are stored before processing. The bloodwater is usually added back to the process to recover the nutrients and to avoid pollution problems (Figure 52). The process usually includes a heating step to 7-80 centigrades to coagulate the protein followed by a centrifuging step to remove the coagulated solids and oils. The resulting liquid is then added back into the stickwater and evaporated. The solid fraction usually goes to the presscake line for drying and the oil, normally dark and of poor quality, is either held separately and sold at discount prices or mixed with the fuel oil and burned.

The recovery of the bloodwater is a major step in improving yields and reducing pollution. The quality of the nutrients in the bloodwater will vary with the quality of the raw material at the time the fish are unloaded and their deterioration during storage in the factory. Based on the limited sampling of the bloodwater in the Paracas plants it is difficult to place an exact value on the recovery of the nutrients. Samples were taken from fresh and spoiled fish. In one case, the solids content of the bloodwater was 60% (spoiled fish) and in others as low as 4% (very fresh fish).

Based on the data that was received, you could expect to lose between 0.5% and 6.2% of the processed blood water as protein and fat depending upon the condition of the fish. Since there was no actual volume measurement of the blood water, it was not possible to relate these figures to the tons of fish landed. The data does indicate however, that once the bloodwater has been coagulated and separated, there is still about 76% of the nutrients in the liquid phase. If the liquid phase is discharged as effluent instead of evaporating it with the stickwater, then the major part of the nutrients are being discarded. The proteins in the bloodwater appear to be soluble and are not being coagulated by the process being used (Figure 53). Evaporation would seem to be the best route to recovering this material.

We've now identified the potential problem areas, defined the challenges and offered options to meet the challenges and improve returns by increasing yields. But what are the estimated costs and the payback to recover the capital investment. It isn't possible here to go into a complex return on investment analysis for Peru, but we can estimate the capital cost and calculate how long it will take to recover that investment with product. Figure 54 outlines the various optional technologies and groups of technology that have been described. In Figure 55 we list the various combinations together with their capital cost, amount of fishmeal that can be recovered, its value and the time it will take to recover that capital. It does not take into account operating costs and maintenance.

11. Other processing methods

11.1. Fish silage

11.2. Composting

11.3. Extruded feeds

11.4. Bio-gas production

11.5. Discharge into the sea

For the small seafood processing plant whose main product is edible fish, the generation of waste streams both solid and liquid present very large problems. If the processing plant does not have or cannot deliver the solid waste (cuttings) to a fishmeal plant, it must make other provisions for disposal of the wastes. Unfortunately, this might involve discharge into the sea or burial in a landfill if one is available. This is a global problem for the small seafood processor. The lucky ones have fishmeal plants that are neighbors and are willing to take the solid waste. The liquid waste from the edible seafood processing plants usually does not contain a large load of nutrients and screening usually removes most of the nutrients in the water.

Fish silage is liquified fish stabilized against bacterial decomposition by an acid. The process involves (Figure 56) mincing of the fish followed by the addition of an acid for preservation. The enzymes in the fish break down the fish proteins into smaller soluble units and acid helps to speed up their activity while preventing bacterial spoilage. Formic, propionic, sulfuric and phosphoric acids have been used. Normally, about 3-4% of acid is added so that the pH remains near 4.0. Strong mineral acids require neutralization before feeding the final product. Silage might be defined as a crude form of hydrolizate.

Silage made from white fish offal does not contain much oil, but when made from fatty fish such as herring it is necessary to remove the oil. The composition of the silage will be very similar to the material from which it is made. Fish silage of the correct acidity is stable at room temperature for at least two years without decomposition. The protein becomes more soluble, and the amount of free fatty acids increases in any fish oil present during storage. Silage production offers a solution to the handling of fish waste, when the logistics of delivering to a fish reduction plant are not economical. Silage can be produced in large or small containers both on the vessel and onshore.

One such operation uses structures (Figure 57) that are 4 feet high x 5 feet wide x 16 feet long using wood and wire fencing. The structures have a 6 inch base of gravel and 8 x 5 foot drainage pipes laid widthwise to provide bottom aeration during the composting process. A 6 inch layer of peat moss was then put down as a base for the compost pile. The peat moss used throughout the pile proved helpful in eliminating odor and rodents. An alternating layered approach was used in setting up the compost piles building 6 inches of fish waste then 6 inches of peat moss. To speed the composting process a commercial compost starter was added along with water to each layer of waste. The completed piles each contained about 5000 pounds of fish waste and cost about US$510 to produce. Wood chips can be used to replace the peat moss and this will reduce the cost considerably. Almost any waste vegetable material can be used.

To control a healthy compost pile you must be aware of the odor (Figure 58 and Figure 59), temperature and composition (Figure 60 and Figure 61).

12. Summary

In many countries where fishmeal production is a major industry, the fishmeal was probably the first permanent installation in the area (Figure 62). Soon, workers move in to be near the source of work and income. The workers are followed by the people who supply services and support for the plant and workers. A town or city develops, more people move in and require more services, schools, medical, living quarters, food and shopping. As the area develops, the population becomes less dependent upon the fishmeal factory and soon there is conflict between the fishmeal factory and its neighbors. People begin to complain about the smell, the noise, the fact that they cannot use the ocean and beaches for recreation. Civilization is squeezing the fishmeal factories. Soon government regulations appear. The plants must meet certain effluent guidelines in order to stay in operation. The plants that are progressive and have the financial means to meet the regulations stay in business, those who cannot eventually are either shut down or close on their own. The regulations usually come in stages, installation of best conventional technology, followed by best available technology. Each time the limitations for effluent discharge are reduced. Primary treatment such as stickwater plants and screening pumpwater are no longer acceptable and secondary treatment must be used. In all cases, the critical point is the amount of solids going out of the plant. Most treatments can't handle the large volumes that are generated, so the industry is forced to reduce the volume to something manageable. Reducing pumpwater volumen, retrofitting evaporators so that condensate water can be separated from cooling water (cooling water can be discharged back to the sea), aerobic and anaerobic lagoons are installed to reduce the discharge load by 90% or more. These are the steps that will eventually come to the Peruvian fishmeal industry. The companies that are prepared to address the issues will survive, the ones that can't or won't will be closed or acquired by the larger ones.

Installing equipment that will recover valuable product and reduce effluent loads, makes good economical and environmental sense. Discharging these materials back to the sea will in the long term hurt the industry because it will demonstrate to their neighbors that they are not concerned with the environment and in today's climate of aggressive environmental groups it could be fatal. Environmental groups now go directly to the consumer, the consumer puts pressure on the supplier of the food products that are sold. The food supplier puts pressure on his supplier and eventually this leads back to the fishmeal plant or other ingredient supplier. This is happening right now.

Figure 1 Peruvian landings Vs fishmeal landings

Figure 2. Estimated losses in the peruvian fishmeal industry 1950-1994

Figure 3. Major key to pollution prevention

Figure 4.

Figure 5.

Figure 06. Key areas to consider in the fishmeal process

Figure 07. When the raw material spoils

Figure 08. Reducing the storage temperature by 5º - 6º C.

Figure 09.

Figure 10. 5 year average (1991-1995) Landings of fish at peruvian ports

Figure 11. Critical assumptions

Figure 12. Metric conversion of MG/L dry solids to KG/TON of fish processed

Figure 13. Calculations to convert MG/L in the water to tons of fishmeal.

Figure 14. Material mass balance for a typical anchovy fishmeal plant

Figure 15. Herring yield loss when stored at different temperatures

Figure 16. Weight losses in variou fish stored at 3 different temperatures over 17 days

Figure 17. Daily losses in weight during bulk storage of fish

Figure 18.

Figure 19. Increase in TVN of stored capelin over time at different temperatures

Figure 20. Change in TVN of herring and small cod over 13 day period

Figure 21. PRESERVE THE QUALITY OF THE RAW MATERIAL

Figure 22. The comulative effect of raw material quality on the fishmeal process

Figure 23. Criteria for ideal unloading systems

Figure 24. Raw material unloading systems

Figure 25.

Figure 26.

Figure 27.

Figure 28.

Figure 29.

Figure 30.

Figure 31.

Figure 32. Unloading and handling systems for fish

Figure 33.

Figure 34.

Figure 35. Pumpwater flow diagram

Figure 36. Wet unloading process disadvantages

Figure 37. Pumpwater losses calculated as fishmeal by port

Figure 38. Dry unloading process disadvantages

Figure 39. Blood water losses calculated as fishmeal by port

Figure 40. Blood water process flow diagram

Figure 41. Stickwater process flow diagram

Figure 42. Loss of solids in the pumpwater with and without screening

Figure 43. Loss of solids and their value from not processing the blood water

Figure 44. Loss of solids in the stickwater if only 25% of the factories discharge the stickwater

Figure 45. Summary of yield and economic losses by processing step

CALCULATED FOR 100KG OF FISH CONTAINING 200 KG DRY SOLIDS PLUS 90 KG FAT, ASSUME FRESH FISH.
IOL YIELD 40 KG PER TON OF FISH OR 4%

Figure 46. Pumpwater challenges

Figure 47. Pumpwater suggestions 1

Figure 48. Recycling of pumpwater

Figure 49. Suggested fish unloading recycle system

Figure 50. Screens

Figure 51. Screen testing results

Figure 52. Blood water recovery

igure 53. Blood water data generated - Pisco plants

Figure 54. Possible technology combination and options

Figure 55. Investment costs and returns by options

¹ For factories with no stickwater plant recover of stickwater only
² Assumes use of the wet pump with fresh water, recycling and additional evaporation
³ Assumes purchase of sufficient pumps, recycle system for fresh water and additional evaporation
⁴ This is very rough estimate since the pump is not being manufactured right now. Ratio water to fish low enough not to require additional evaporation
⁵ No water used, includes dry unloading, conveyors and pier

Figure 56. Flow diagram of the silage production process

Figure 57. Simpe fish waste composting system

Figure 58. A sick compost pile 1

Figure 59. A sick compost pile 2

Figure 60. A sick compost pile 3

Figure 61. A sick compost pile 4

Figure 62. The environmental dilema