Evaluation of Fish Waste Management Techniques
2 Description of Available Fish Waste Technologies
At present the options for storing, utilising or disposing of fish wastes are limited to the following: (depending on whether they are Category 2 or 3)
- Freezing: storage of material for use in pet food manufacture etc.
- Ensiling: essentially a storage technique for subsequent disposal or utilisation
- Rendering with other material: reduction of material to fish oil and bone meal
- Reduction to fish meal and fish oil: conversion to a marketable commodity
- Direct consumption: utilisation by zoo and circus animals, hounds, maggot and worm (for use as bait) farming
- Incineration: burning of de-watered material
- Landfill: restricted to remote areas where no alternative solutions are available
The following section provides a detailed breakdown of these current options, including their costs and regulatory restrictions. At the end of each sub-section is a 'Strengths, Weaknesses, Opportunities and Threats' (SWOT) analysis. A comparative analysis of these and emerging technologies is contained in Section 4.
Where possible capital, revenue and value-added costs are provided for each technology. The basis of these costs, together with a summary, is provided in Appendix D.
In the 1980's the freezing of fish waste destined for the pet food and fur trade was an important outlet for waste from the fish processing industry. Now only a small tonnage of waste ( c. 2,000 mt) is plate-frozen in Hull and Dumfries and sold via traders to pet food companies in the UK. The decline in this pet food market has resulted mainly from a move from canned products to pouched and extruded feeds. The purchasing department of Britain's largest pet food manufacturer confirmed that the move away from canned into dry and pouched pet foods will limit the use of frozen salmon waste for pet foods. They added that white fish offal was preferred for visual reasons in the majority of recipes and that fish offal use was unlikely to grow as odour issues limited its use overall, no matter the superiority of the nutritional value. 25% of frozen material is also exported to the Scandinavian fur farming trade now that fur farming has ceased in the UK.
The plants mentioned above are the only two dedicated to pet food and mink feed. However the technology concerned, vertical plate freezing, is widely available throughout the fish processing industry. It is used to freeze large seasonal supplies of fish, mainly pelagic, in order to iron out disparities between filleting capacity and spot supply and for export. It is understood that the Scottish fish processing industry currently has excess freezing capacity should the demand for frozen wastes increase.
Using FAO Technical Paper 340 as a guide, and assuming that waste material is frozen in a horizontal plate freezer of one tonne per hour capacity running for 4,500 hours per year, the following costs are estimated in Table 6 overleaf.
Table 6: Freezing costs (1 mt/hr capacity)
Depreciation (10 years)
Other fixed costs
Total fixed costs
Total variable costs
Cost per tonne
Value added per tonne
SWOT 1: Freezing
- Suitable for all forms of material
- Can be used to store material to ease seasonal gluts
- Restricted to Category 3 material
- High capital costs for freezing and storage facilities
- Expensive process
- High storage costs
- Requires cold chain
- Mink farming now only outside UK
- Greater utilisation of spare freezing capacity
- Opportunity for monopoly / market leader
- Declining market for fish waste-based pet foods
- Competition from abroad
2.2.1 The Process
Fish silage is an aqueous product that can be produced from any species of fish or fish waste. In Scandinavia white fish waste has been regularly ensiled from the period prior to the establishment of the salmon farming industry. The production of fish silage is a relatively simple, low technology process. Its simplicity removes the basic minimum economic size constraint making it ideal for small waste flows or those situated in remote areas. Ensiling does not require high energy inputs, odour or emission control of any sophistication nor any pre process preparation. Supplementary materials are limited to formic acid.
Ensiling units can range in size from 250 L to 2000 L though in practice smaller sizes prevail. In contrast to fishmeal production it is conducted on a widespread basis at a local farm and processor level. The majority of ensilers are situated at fish farm land-base sites to deal with 'routine' mortalities and minor epizootic events. Because they are mobile (although in practice most are never moved), ensiling units can be located close to the source of the raw material thus eliminating the necessity for temperature controlled storage and transportation facilities. It is important with ensiling, as with all recovery processes, to aim for the freshest raw material possible. Stale or rotten fish inhibit the speed of the process. The end product has similar advantages, needing only to be kept in sealed containers, which can range from 50L drums for morts to 30 tonne tanks or silos for process waste.
The process has few stages. In the first the fish passes through a macerator. The resulting increase in surface area makes the enzymes in the viscera more available and speeds the liquefaction process. Without viscera the process takes longer. The next stage consists of pumping the macerated raw material to a mixing tank where it is mixed with formic acid at 3.5% to acidify it. The pH is kept at 4 or below which allows the fish offal to autolyze without spoiling. Finally the silage is pumped to a storage container. Where processing waste is involved, tanks, typically of 15 to 30 tonnes capacity, are commonly used. Mortalities are more normally held in 50L drums.
Some units have the ability to further separate the fish oil, but most simply hold the material until it can be collected and either disposed of or further processed. Removal of excess oil provides revenue in itself and is necessary to enable the silage to be used for some applications such as spreading on land as a fertiliser (Category 3 wastes only). Excess oil can be washed into watercourses with discharge problems arising as a result.
To remove the oil it is necessary to heat the silage, then skim off and further centrifuge the oil to clarify it. At least one trout farmer does this but unfortunately the scale of the operation (>150T per year), means that the equipment lacks the sophistication to produce oil of high quality and it can only be used as a fuel or as feedstock for biodiesel.
Ensiling is also used to deal with water/blood water treatment plant sludge from those processors so equipped. This source of material follows a different disposal route from that being discussed
Silage can also be produced by adding a carbohydrate source such as simple sugars or molasses. The process requires an optimum temperature of 25 -30 °C and so is restricted to use in climates that can maintain that temperature without supplementary heating and is not further considered here.
2.2.2 Volumes and Extent of Use
It has been estimated that less than 15,000 tonnes of waste from aquaculture mortalities and primary processing waste is currently ensiled in Scotland. As far as we are aware, apart from English trout farms and Irish salmon processors and farms, no ensiling of primary wastes takes place elsewhere in the British Isles. The volumes of material uncollected or uncollectable by the other technologies mentioned in this section are probably not much greater than this.
Ensiling is recognised as a method of pathogen inactivation and is recommended as such in the ISA code of practice (Anon, 2000).
The majority (4) of primary processing plants in Shetland and Harris ensile viscera and at least in one case, frames as well. Some mainland units have ensiled in the past but at present, in common with the majority of inland plants, have their process waste removed for reduction to fishmeal.
In Scotland four companies compete for ensiled material. One, Rossyew Ltd based in Greenock, collects and processes viscera and then only Category 3 material. It has just recently commenced operations. Installed capacity is sufficient to handle 20,000T of material a year.
Another Scottish operation - Emac - based in Invergordon collects blood water/water treatment processing sludges from a number of processors. These have been injected into farmland after heat treatment under an exemption to the Waste Management Licensing Regulations but this has been rescinded. The volumes are small, less than 1000 mt pa. Emac is a waste collector and processor whose mainstream business is dealing with distillery waste. They also collect a substantial proportion of farm mortalities, which they transport onwards for disposal by incineration in England. They are understood to be developing a process which would eliminate this costly second step but no details are forthcoming at the time of publication of this report.
Two Norwegian companies, Scanbio and Hordafor also provide that service. Scanbio has storage facilities in Inverness from where several thousand tonnes of Category 2 and 3 materials are shipped to their parent company for further processing and sale. This means that much of the potential value of the process waste is lost. Possibly this is because there are insufficient amounts of each Category to permit separate collection and processing. Hordafor has temporary storage capacity in Stornoway Harbour and last year announced its intention to erect a processing facility in the disused oilrig construction yard at Arnish. Whether sufficient material is available to justify such a venture is discussed later. The same company announced a similar venture in 2000 at Lerwick (March 2000, Shetland Fishing News) which failed to materialise.
In Norway Scanbio and Hordafor dominate the business. Both utilise salmon and rainbow trout offal and white fish offal from the traditional processing industry. We have as yet no figures for the latter sector but understand they are equivalent to the amounts coming from salmon processing which are in excess of 90,000 mt per annum. Hordafor claim in their literature to process in excess of 80,000 mt annually. Scanbio's output is a little smaller. This gives the industry in Norway a critical mass that is unlikely to be achieved in Scotland.
2.2.3 Outputs and Disposal
As a result of the ABP Regulations, from 1 st January 2004 Category 2 ensiled material -predominately from morts - will have to be either further heat treated prior to its use as a fertiliser or incinerated. This latter route was the one adopted for the large fish kills in Inver Bay in Ireland last summer where over 400,000 fish were lost. The fish were rendered first at Ballynasloe near Athlone then transported to Germany for incineration, there being no suitable incinerators in Ireland. The enormous costs involved are probably less than the costs of using dedicated units with virtually no regular throughput.
Possibly only 50% of silage from mortalities in Scotland has been further processed abroad to date, the remainder being land filled or stored whilst waiting disposal opportunities.
Category 3 ensiled materials will still be permitted as an ingredient in animal feeds after appropriate heat treatment. Unfortunately liquid raw materials have limited application in the UK compounding trade. Firstly compounders have to comply with the fishmeal ban and so silage use would be restricted to mills dedicated to non-ruminants. As the inclusion of silage would be only be up to 10% of the ration and as their process is dry pelleting rather than extrusion, their reluctance to invest in new injection technology and storage for the wet silage is understandable.
Silage from Category 3 waste is a widely established product in Scandinavia, Poland, Denmark and Holland. It is highly nutritious and is fed to pigs, fur animals, and poultry as an ingredient in diets for these animals. Some is injected into soil as a natural fertiliser. Silage from mortalities and other Category 2 material is shipped to Denmark as biogas feedstock. Silage from white fish is used, as an alternative to LT fishmeal, by Skretting, Norway's largest fish feed producer and sells at a slight discount to that product on an equivalent dry matter basis. (LT meal today is 355/mt FOB Peru). Some producers have developed the product further by reducing the water content. The savings in freight thus obtained how ever tend to be offset by higher energy costs and fully dried silage cannot be differentiated from fishmeal.
Unlike their Norwegian counterparts who get paid for the raw material, Scottish fish farmers face costs (gate fees) averaging 38/mt for collection of offal. Apart from the set up costs for the ensiling equipment and the variable operating cost of formic acid, the fish farmer or processor faces a confusing scenario in Scotland as to downstream costs. Three, possibly four collectors and further processors of silage compete for 15,000 tonnes of material. Compare this with the situation as described previously in Norway. Since this supply comprises both Category 2 and 3 materials as defined in the ABP Regulations, then unless the operator has fully segregated facilities, Category 3 material has to be downgraded to Category 2. This means that it cannot enter the feed chain, has to be further treated at considerable expense, has lower prospects for revenue production and introduces diseconomies of scale to the operation in question.
Although the ensiling process is not complicated, there are significant variations in:
- scale of operation;
- degree of automation;
- nature of raw material.
These factors generate a wide variation in the cost of ensiling a tonne of raw material. The table below compares an automated operation using a 2,000 litre ensiling machine with an example of ad hoc production, both assumed to take place at a sea cage site producing about 1,000 tonnes per year.
Table 7: Costings for a 2000 L Capacity Ensiling Plant
Ad Hoc Operation
Primary processing (viscera), tonnes
Total waste, tonnes
Ensiler, 2,000 litres
10 x 1,000 litre IBCs at 80 each
Ad hoc equipment (macerator, pump etc)
Ad hoc tank
Total capital costs
Ensiling equipment (10 years)
Storage facilities (15 years)
Formic acid, 3.5% for mortalities, kg
Formic acid, 2.5% for viscera, kg
Total formic acid at 150 per 240kg
Labour at 4.50 per hour
Total cost using 3.5% acid throughout
The main difference in costs between automatic and ad hoc ensiling arises from the labour saved by automation. The labour input of one hour per tonne is quoted as the maximum for automatic ensiling. The four hours per tonne for ad hoc production is from FAO Circular 905 which reports that a single operator working in this way would produce two tonnes of silage per shift. One aquaculture site which provided ad hoc capital costs, estimated labour at five hours per tonne.
The difference in capital costs arises mainly from different approaches to storage. With the automatic ensiler it is assumed that individual bulk containers would be used, for convenience of handling and transport as well as storage. The remaining factor affecting costs is the percentage of formic acid needed. Less is required for treating viscera than for mortalities or secondary processing waste (heads and frames). The final row in the table above shows total costs using formic acid at 3.5% throughout.
For the purpose of this study, ensiling costs will be estimated on the basis of automatic ensiling machines, about 28/tonne at a sea cage site and 32/tonne for a secondary processor. A suitable standard cost for ensiling will be taken as the average of these, i.e. 30/tonne.
Land Injection of Ensiled Material
Although landowners may be presumed to benefit from treatment of their land, reports suggest that the only payment to the service provider is that made by the waste producer. The charge for land injection is attributed mostly to transport, but no breakdown between transport and injection itself is available. A typical charge for land injection is 25 per tonne of silage, of which 80% is here assumed to be for transport. Adding the cost of ensiling, this brings the total cost of disposal by this method to around 55 per tonne.
Value added: Given that the land injection service consists mainly of transport, the ratio between wages and profit components of value added is assumed to be the same as already estimated for carrying IBCs from the west coast to Inverness.
Table 8: Value Added - Land Injection of Ensiled Material
Estimated value added, /tonne
land injection of ensiled material
Transport and Export
Two Norwegian firms are presently active in collecting and shipping Scottish fish waste to Norway for processing. One, Scanbio, collects only ensiled material. The other, Hordafor, collects both raw and ensiled waste.
Scanbio costs and value added
Since 1998 Scanbio Scotland has provided a silage holding facility of capacity 2,000 mt at Inverness, charging producers 38 per tonne for the service. Silage is collected at three to six month intervals by a 1,200 tonne vessel for shipping to Norway. Producers pay transport to Inverness in addition to their own ensiling costs, giving a total cost of around 89 per tonne for this method of disposal, assuming a 150 mile trip. Scanbio Scotland employs two persons. Assuming average wages of 6.00 per hour and an annual throughput of 5,000 tonnes, the indicated value added would be as follows.
Table 9: Value-added (/mt) for Scanbio Scotland
Estimated value added, /tonne
With throughput in 2003 down to about 3,000 tonnes, profitability was reduced so that value added is now probably less than 8 per tonne.
Hordafor costs and value added
Shetland packing stations located on the harbours at Lerwick and Scalloway, under agreement with Hordafor, ensile all process waste plus approved mortalities from farms, storing the silage in intermediate bulk containers (IBCs). Every three weeks a Hordafor vessel arrives from Norway to take it away free of charge. For most of the waste, say 75%, the only cost to producers is that of ensiling, 30 per tonne. The remaining 25%, representing viscera and approved mortalities, bear an additional small transport cost, probably around 5 per tonne. The overall average cost to producers is thus estimated at about 31 per tonne. The only value added arises from the transport and from periodic quayside labour transferring silage to the ship, probably less than 1 per tonne overall.
At Stornoway, Hordafor have stationed one of their vessels, with a capacity of about 500 tonnes, in the harbour to provide a variety of services. The single producer so far contracted to Hordafor sends secondary processing waste (heads and frames) plus approved mortalities directly to the ship for ensiling on board at Hordafor's expense. Viscera and the remaining mortalities are ensiled by the producer before being sent to the ship where material can be stored separately according to quality in the vessel's eight tanks. The ship also acts as a standby in case of a catastrophic mortality event.
Assuming that Hordafor's Western Isles partner has to ensile only 20% of waste, the average cost to this producer for ensiling would be 6 per tonne overall. This producer has enough spare transport capacity for waste to be taken the short distance to the quay at negligible cost. For other potential partners on Lewis at least, the cost of silage transport to Stornoway would be small, probably under 10 per tonne using IBCs. Any Western Isles producer working with Hordafor could make large cost savings by not having to use the very expensive local landfill sites.
With the only remaining expense the hire of a forklift truck at the quayside, the value added specifically from using Hordafor's Stornoway facility is probably no more than 1 per tonne. Additional value added is derived from Hordafor's purchases of local goods and services, plus whatever Hordafor crew members spend in Stornoway, but these are ignored here.
Hordafor's original plan to build a new factory at Arnish for processing silage (1 st grade for pharmaceuticals and cosmetics, 2 nd grade for high protein animal feeds, 3 rd grade for biogas) has been delayed, but it is reported that to be profitable such a venture would need to draw upon raw material from Shetland and Ireland as well as the Western Isles. Although some of the wages and profits would benefit Norway, the remainder of value added could accrue to Scotland.
SWOT 2: Ensiling
- Suitable for most types of finfish waste material
- Simple, low technology process
- Can operate at a wide number of scales from 250 L to 2,000 L capacity
- Low energy and other input needs
- Low odour and other emissions
- Essentially a treatment and stabilisation mechanism - ensiled material still needs disposal
- Requires only fresh material (cannot deal with putrefied waste)
- High oil content of salmon wastes may necessitate further separation.
- High cost of collection and transport.
- Category 2 material requires expensive heat treatment to allow its use
- Category 3 material downgraded to Category 2 if not separated.
- Products can only be used for non-ruminants.
- Competition from abroad
- Suitable for Category 2 and 3 materials
- Widely used in Denmark as a pig feed if from Category 3 material
- Development as a semi-moist feed or paste product to reduce transport costs and provide new markets.
- Further process has export potential e.g. for shrimp culture
- Low value of ensiled product requires high volumes that may not be economically possible in Scotland.
- Subject to dioxin restrictions.
2.3 Reduction to Fishmeal and Fish Oil
The traditional manufacture of fish meal involves some or all of the following stages.
- Mincing of the raw material (optional). Most plants can accept whole fish.
- Heating of the waste between 82 and 90° C in a cooker for 15-20 minutes partly to ensure the destruction of both pathogenic and spoilage organisms and to facilitate the next stage.
- Pressing, which separates the solids (press cake) from the liquids (press liquor) containing oil and water
- The press liquor is then decanted to separate the water and oil from the solids.
- The decanted liquor is centrifuged to separate the water and oil.
- Next the oil is stored and allowed to cool
- The water is evaporated to retrieve any soluble solids and these and the solids from decanting are re-combined with the press cake, which is then dried.
There exist a wide variety of drying technologies. Older processes that can be direct contact dryers or direct flame dryers operate at 90-95°C whereas the more recent types, for so-called 'low temperature' (LT) meal, operate at 60-65°C. These dryers predominately use indirect air heating. These are the temperatures of the meals in process, not the drying medium temperatures that can be considerably higher.
Grinding the meal completes the process. It is then either bagged or, more often, sold in bulk. The final product typically contains 65-70% protein, 8-10% moisture, and 10-12% oil, the balance being ash, salt and sand. These proportions are dependent on whether whole fish or process trimmings are used, the species involved, the freshness of the material as measured by TVN, temperature and the characteristics of the plant used. It is considered to be sterile or nearly so by virtue of the initial cooking process. This does not mean that pathogens do not occur in fishmeal but when they do it is usually as a result of contamination by rodents or birds. The upper limit of 12% fat is necessary to prevent spontaneous combustion (even with antioxidants).
Fish meal is used in most types of animal feeds at levels typically of up to 45% in fish feeds. 5-10% in pig feeds and 1-5% in poultry feeds. Aquaculture took 55% of the 2002 UK consumption of 235,000 mt, 45,000 mt of which was produced domestically. Imported meals, mainly from Peru and Scandinavia come in a much wider range of types and qualities than UK meals. Despite this, the published price CIF Hamburg (Oil World - Mielke Hamburg) reflects most transactions and insofar as a market can, fishmeal operates as a near perfect one. The major external factor on fishmeal prices is the effect of soya as a substitute protein in Far East aqua feeds and animal feeds. Unlike soya and other protein sources fish meal supply is static and controlled by governmental fishery policies. Hence when events like an 'El Niño' occur supply can be halved and prices double. At present the price is 425 ex store (UK)/tonne.
Demand for fishmeal in the UK has dropped considerably from past levels (IFFO annual statistical yearbook) of 284,000 mt in 1998 to 193,000 last year. There is currently in place a ban on fishmeal use in ruminant feeds in the EU which accounts for half this drop but a 15% decline in the UK pig herd and changes in husbandry practices in the poultry industry have also contributed to the drop in demand from these sectors. The total amount incorporated in aquaculture feeds fell for the first time in 12 years in 2003 and will drop further until that sector recovers from its current malaise. Nevertheless worldwide demand for fishmeal is extremely buoyant and so prices are at high levels historically and are expected to remain so. The ruminant ban had no discernable effect on prices, which are set by world demand, especially that from China and aquaculture. Lifting of the ban is not expected to affect prices either.
One tonne of fish waste produces between 2 to 8% of fish oil after processing. The yield depends on the species of fish, whether pelagic like mackerel or demersal like cod, seasonality, the condition of the fish, and whether whole or offal. Fish oil is a valuable by-product from fish waste. During 2003 the market price CIF northern European port (any origin) has risen to $US 610, close to the historical peak reached in 1998 during the last El Niño event. The weakening of the dollar against both the and the Euro has kept the price stable in Europe. 95% of oil produced domestically from traditional non-aqua sources is ends up in fish feed as does 65% of world production. Most commentators agree that all but a few percent will be consumed by aquaculture within 5-7 years, forcing substitution by vegetable oils in aqua diets ( see Figure 3). This has been done before but on cost grounds. Now that, for the present at least, soya and rapeseed oils are more expensive than fish oil, substitution will be out of necessity not economic choice. Should the level of substitution be limited to, say 25% of the fat content of an aquaculture diet then this limited availability of fish oil will become the limiting factor for growth in carnivorous aquaculture species production.
When fish oil is obtained from the ensiling process the oil so produced has a slightly higher FFA (Free Fatty Acid) content than that from the fishmeal process given raw material of equal freshness. Usually the levels are such that it is unsuitable for aquaculture but acceptable for other animal feeds. Where it is a by-product from aquaculture waste then, as mentioned it cannot be used for aquaculture, or, at least in the EU, for farmed fish 5. A market for salmon oil exists in the human health industry, for pet food and in aquaculture in Japan and the Far East. With 30,000 tonnes coming from Norway and 8000 tonnes from Chile supply is more than the market can absorb at present. Consequently salmon oil sells at $100/ mt below conventional fish oil. In Europe it goes to the 'technical' market which includes leather tanning, special lubricants, health products etc.
Figure 4: Global Fish Meal and Fish Oil Usage (2002 and Predicted 2010)
Source: Barlow (2002)
Of an estimated 162,000 mt of fish waste generated in the Scottish capture fisheries ( see Figure 1) the majority, over 70%, is reduced to fishmeal and fish oil. The conversion yield is approximately 20%. This activity is primarily concentrated in a few operations based in major fishing and processing centres like Grimsby and Aberdeen. Shetland has a medium sized plant and there are three other very small plants in East Linton, Grimsby and Plymouth. The latter is a modified pilot plant from 999 (formed from a merger of Esbjerg Fiskindustrie and Thyboren Andels) in Esbjerg in Denmark.
The factory in Aberdeen has a capacity of 45 mt/hour and occupies a site of 4 acres. Much of this is occupied with storage silos for incoming raw material, oil storage tanks and fishmeal warehousing. The smallest plant, in Plymouth, processes only the mackerel trimmings from an adjacent freezing and processing unit and is remarkably compact with a footprint of 500 m 2. Storage on site is limited to 30 tonnes of oil and 30 tonnes of meal so frequent collection and third party storage of the finished product are called for.
All UK plants are licensed to deal with Category 3 material only, under the ABP Regulations. This means that except in certain exceptional circumstances, they cannot process morts or casualties from aquaculture. All these units are in, or close to, areas traditionally associated with the fishing and fish processing industry and are the remnant of a once much larger and more widespread industry. This is the typical model in the majority of countries with a sizeable fish processing activity. Peru is possibly the only exception to this model having experienced growth and investment for some years past.
Although the bulk of world fishmeal supplies derive from fisheries dedicated to that purpose, process wastes are responsible for 11% of world supplies. (FAO, 2002) One exception, surprisingly, is Alaska, (Bethel, 2002) where much process waste is taken back out to sea, minced and dumped with what many view as serious environmental effects on the very fisheries from which the material arose (Bloom, 2002).
Fishmeal plants can accept almost all kinds of aquatic inputs. Fish size and type can range from large whole fish like cod to small fish such as anchovies. Offal of most kinds can be accepted including fish frames and viscera, the only physical limiting factors are the free water content and the total water content. Thus viscera can only be processed with an equal amount of solid material such as frames. Shellfish viscera from scallops are almost unprocurable. Crustacean and shellfish offal is not normally dealt with in the plants mentioned because its' low protein but high ash content cause processing problems and render the resulting meals unsuitable for many end user formulations. For example heavy use of high ash meals in fresh water trout diets means that farmers cannot meet their water discharge permit levels.
The other inputs are considerable amounts of power, some enzymes and water. There are no pre-input treatments required. The condition of the raw materials dictates final product quality and thus haulage, storage and process conditions are strictly controlled. The units mentioned have a collective capacity on a year round basis of 800,000 Mt per year. Only 250,000 Mt of this is used due to the seasonal nature of supply e.g. demurral supply is greatest in winter, pelagic inputs are brief but heavy, and a general decline in overall supply since the plants were built.
The aquaculture industry imports shrimp-meals from Iceland and Chile each year as local supply has up to now been non-existent. Moray Seafoods have recently commenced to produce a prawn meal for aquaculture. The process is not strictly a meal process as most of the stages described previously are omitted. As the raw material is already cooked, there is no cooking stage, no oil to be removed and the process is simply de-watering, drying and grinding. One other major prawn and seafood processor has built a plant to deal with their waste but, as yet, has not gone into production citing technical reasons.
Product yields and prices
The average yields from secondary processing salmon waste reduced to fish meal and oil at a plant such as the one at Grimsby are estimated at 20% meal and 5% oil. Estimates for medium to long term product prices are, per tonne: US$580 for meal; US$600 for oil. At the present exchange rate these convert to about 330 and 340. Thus, one tonne of raw material is estimated to produce by value: 65.91 of meal; 17.05 of oil; total 82.95.
Costs and value added
The approximate capital cost to build a fish meal plant capable of processing from 50 to 100 tonnes of raw material per day is about US$ 15 million (8.5 million). This excludes investments necessary to supply raw material (e.g. boats and vehicles). European fish meal producers are generally prepared to pay between 35 and 40 per tonne of raw material. One producer interviewed as part of this study collects and transports fish processing waste free of charge. As already described, transport of raw salmon waste from Fort William to Grimsby costs about 32 per tonne, which can be taken as the raw material cost.
Fixed wages for a fish meal plant of 50 to 100 tonnes per day capacity amount to approximately 450,000 per year. In this capital-intensive process, variable wage costs are negligible. Thus, a plant of this size processing 100,000 tonnes of raw material annually would spend 4.50 on wages per tonne of raw material.
With major fluctuations in raw material supply, international product prices and exchange rates it would be optimistic to assume a profit after depreciation at 10% of turnover. For the purpose of this study, profit for a fish meal operation is assumed to be 6% of turnover. Under the stated assumptions, estimated value added would be as follows.
Table 10: Value Added for Fish Meal and Fish Oil Production
Estimated value added, /tonne
Fish meal & oil production
SWOT 3: Reduction to Fish Meal and Fish Oil
- Increasing domestic and global demand for fish meal and oil, esp. from aquaculture
- Accepts a wide range of materials
- Stable and high value products
- Existing collection and processing infrastructure
- Limited to Category 3 material
- Only wild-caught fish wastes can be used for aquaculture feeds
- Largely appropriate for large-scale industrial processing
- High energy requirements
- Will continue to be the main waste outlet for capture fisheries wastes.
- Landing of material currently dumped at sea, possible subject to a future discarding ban.
- Increased value-adding opportunities.
- Development of farming of marine species such as cod and haddock may introduce traceability issues.
- Price-competition from other plant-based protein sources
- Volatile prices - susceptible to supply of targeted industrial feed fisheries e.g. South American anchovy
- Further restrictions in the use of fish meal in ruminant diets.
- Dioxin levels in raw materials, salmon as well as wild catch and concurrence with EC legislated limits.
2.4 Direct Consumption
The Animal By-Products regulations permit the use of Category 2 and Category 3 material for use in processes not destined for human consumption such as feeding to zoo and circus animals, fur animals (now not permitted in the UK), maggot farming and vermiculture.
Less than 500 mt per annum is estimated is taken fresh and unprocessed, in the summer only, by maggot farmers. The operators of these units are prepared to pay as much as 60/ mt collected for material. None of these operations can properly be described as being a technology since the process consists of raising larvae in a growing medium in plastic containers. The operation is only viable when ambient temperatures allow, usually May to September. Suppliers complain of erratic collection, late payment and unpredictability. The possibility for maggots to act as a disease vector is likely to discourage this application, at least in Scotland.
Whether any fish waste is being used directly in vermiculture is not known and is presumed to be unlikely and though anecdotal accounts suggest that worms will break mortalities down quite rapidly (source FRS staff), other routes would seem to have more potential for treating volumes of waste without any associated problems with disease. Seafish will be supplying material from their composting trials for vermiculture in May 2004.
SWOT 4: Direct Consumption (non-human)
- Limited, but high value demand for suitable raw material
- Limited to Category 2 and Category 3 materials
- Limited demand from zoos
- Logistics of transporting and handling fresh wastes
- Seasonal demand (mainly summer)
- Vermiculture increasing in popularity.
- Maggot farms are often small operations with unreliable demand
- Vermiculture and maggot farming have potential for the recirculation of disease
Some waste material from farmed Scottish fish is presently incinerated at two locations, one in Shetland, the other at Widnes. The Shetland Energy Recovery Plant accepts raw mortalities waste for incineration in the conventional way. Installed capacity is 80 tonnes per day. Average throughput in 2003 has been about 65 tonnes per day, but very little of this has been fish waste - only about 12 tonnes for the whole year. The management are not happy to take large volumes of fish, as the calorific value is high, especially when the waste contains plastics.
The PDM GroupWidnes fluidised bed combustor, which started operation in 2000, uses fluidised bed technology to co-incinerate meat and bone meal with various waste liquids. The liquids are needed to make a suitable paste for input, but they also enable the plant to produce up to 12 tonnes per hour of high pressure steam used elsewhere on the site. One of the liquids used is fish silage (others include reject soup, blood, etc). In 2003 input per week has averaged around 1,000 tonnes meat and bone meal with around 1,200 tonnes of liquids. Intake of fish silage in 2003 is estimated to be usually between 50 and 70 tonnes per week (min 25, max 100 mt). This suggests a fish silage throughput of about 3,000 tonnes for 2003, some of which is reported to originate as processing waste, not just mortalities. The plant employs a total of 18 persons.
2.5.1 Cost of Incineration
At the Shetland Energy Recovery Plant the gate fee for raw fish waste is 23.07 per tonne. At the PDM Group Widnes fluidised bed combustor the gate fee for fish silage varies between 30 and 35 per tonne, depending on the solids content. For the purpose of this study a gate fee of 32.50 is assumed. The total cost of disposal by this method including ensiling and transport, is around 105 per tonne.
The capital cost of the Widnes plant is not available, but a conventional incinerator of the same capacity and capable of accepting raw material covering the same range of calorific value would cost between 12 million and 15 million. For the purpose of this study it is assumed that the capital cost of a fluidised bed combustor such as the one at Widnes is 18 million.
2.5.2 Value added in incineration
The only significant value added generated by incineration occurs at Widnes. Possible revenue from its output of steam is ignored here. Average wages of 7.00 per hour and annual raw material throughput of 110,000 tonnes are assumed.
Table 11: Value Added for Incineration (PDM Group Widnes)
Estimated value added, /tonne
incineration at PDM Group Widnes
SWOT 5: Incineration
- Suitable for Category 1, 2 and 3 materials
- Fish wastes are difficult to incinerate due to the high water content.
- Residual wastes have to be disposed of.
- Limited options for cost recovery
- High cost - around 105/tonne including ensiling, transport and incineration
- Mainly option for mortality disposal
- Decreased end waste in waste stream
- Increasing emissions legislation
2.6 Land Fill
The balance of the 190,000 mt of seafood waste material originating in Scotland not being dealt with but one of the methods described previously, has been either deemed unsuitable for conversion to fishmeal (usually shellfish because of high ash and low protein levels) or difficult to collect economically (because of remoteness from processing points) or, as aquaculture mortalities, forbidden, and so has up until now gone for landfill. Minuscule amounts have been received by renderers.
The enactment of the ABP Regulations means that land fill is no longer available, at least not without an intervening process such as rendering. Furthermore this legislation, by classifying mortalities from ALL epizootic events as Category 2 material whether caused by disease or by accidental means such as asphyxiation or red tides, prevents it being dealt with by Category 3 plants as in the past. As there are no Category 2 plants for fish at any capacity level in Scotland, then, at present no route is available to deal either with such events or with the mortalities and ensiled material which was previously landfilled.
Derogations are in place which will allow landfill in "remote areas". This is defined as all of Scotland above the Great Glen ( see Appendix C). Annex VII chapter 1 of the ABP Regulations allows for some discretion where facilities (Category 2) do not exist to deal with epizootic outbreaks. SEERAD have stated that in exceptional circumstances they would be prepared temporarily to re-classify a Category 3 facility as a Category 2 one providing a number of conditions were met but this sympathetic approach by SEERAD in the transitional period is not going to provide a longer term/permanent solution to these issues.
Although landfill cannot remain as a general option for disposal of mortalities waste, it has provided a benchmark against which to judge evolving costs of more recent alternatives. The total cost including transport and landfill tax (April 2004) is reported to lie usually in the range 100 to 150 per tonne. The examples in Table 12 below illustrate the point: cheaper landfill sites are often accessible only by expensive journeys; some others close to farm sites have had to be used despite high gate fees when there was no alternative means of disposal. The table below relates to raw waste. Ensiled waste would cost the producer about 30 per tonne more.
The fees quoted are used to illustrate the range of charges encountered, not to suggest what the overall average might be. Any average derived from this small sample can be expected to deviate somewhat from the true mean in either a higher or lower direction.
In the future it is expected that landfill tax rates will increase much more quickly than hitherto, whilst many existing landfills will be shut down through falling short of standards prescribed by new regulations. Therefore landfill costs will rise steeply, making this an even less attractive option.
Table 12: Landfill Associated Costs
/tonne fish waste material
gate fee, landfill
Peterhead, from W coast
SWOT 6: Land Fill
- Lack of processing needs
- Suitable for remote locations where no alternative disposal options exist.
- ABP Regulations increasingly restrictive on what material can be used.
- Environmental threats from liquefied land filled biological material.
- High gate fees due to lack of alternative disposal options
- Limited, given EC/Government policy to reduce land fill
- Increasing costs e.g. land fill taxes increase
- Compliance with EC legislation e.g. leachate issues
2.7 Fish waste treatment in other countries
Norway is an interesting example as the salmon farming industry is well developed here and there are similar logistical, geographic and demographic characteristics to Scotland. As might be expected from the nation who are the dominant force in aquaculture and aquaculture research, Norway has well established routes for dealing with all kinds of fish waste often at considerable volumes. A pioneer of Norwegian waste treatment technologies is the RUBIN foundation, which works for increased and more profitable utilisation of by-products from the fisheries and fish farming in Norway. Several ministries, the Norwegian Research Council and the Norwegian fisheries and industry founded RUBIN in 1992. Since 1998 the fisheries and industry (included fish farming), represented by the Norwegian Fishermen's Association (NF) and the Norwegian Seafood Association (FHL) are the 'owners' of the foundation. RUBIN is financed partly by the Ministry of Fisheries and FHF (Norwegian Fisheries and Aquaculture Research Fund). The board of the RUBIN foundation consist of two representatives each from NF and FHL. No comparable operation has been identified in any other country with an aquaculture industry. RUBIN were instrumental in developing a thermophilic fermentation process (described in Section 3.1)
Norway has a large fishmeal industry which does not deal with any aquaculture by-products, only capture fisheries waste. Output in 2002 was 240,000 tonnes of fish meal and 62,000 tonnes of fish oil derived from 1.2 million tonnes of wild fish.
As described in Section 2.2, two sizeable companies collect the majority of Norway's salmon and trout waste. A number of small businesses operate in the pharmaceutical, cosmetics and fine chemicals areas. One such business produces DNA and fish gelatine (as a replacement for mammalian gelatine) stimulated by the BSE situation with both Roche and BASF using this as a carrier and protective cover for astaxanthin pigment for salmon. This application has to be derived from materials of non aquaculture origin. Hence it benefits only the traditional capture sector.
In 2004 Marine Bio-products' fish protein hydrolysate business should commence operation. Founded by ex-employees of Hordafor and heavily dependent on work conducted by RUBIN, this will provide a further outlet for Norwegian aquaculture waste.
On a cautionary note however Silfas, Norway's second largest fishmeal manufacturer, recently diversified into producing similar products for human consumption applications, including hydrolysates from fresh salmon waste. As Silfas has just gone into receivership this may prejudice the future for Seagarden, as their new venture was named. Technical problems and lower than expected sales are said to have contributed to losses at Seagarden which in turn, compounded the parent's problems.
France has a small emerging sea bass industry but at present of its aquaculture production of 59,200 Mt (2002 FEAP), 42,900 Mt (83%) is rainbow trout in freshwater. According to the secretary of the Association de Transformateurs de Truite (ATT), farm morts are either (i) ensiled and then incinerated or (ii) treated at the fish meal plants in Concarneau and Boulogne sur Mer then incinerated. Processing waste also goes to these plants where it is treated separately. IFREMER is engaged on programmes to produce more value from waste than the cost of disposal but at time of going to print details of these had not been established.
Ireland produces circa 20,000 mt of salmon annually. Unlike Scotland, where farmed fish are not the concern of the Seafish Industry Authority, they are dealt with in an integrated fashion together with capture fishing under the BIM. A 'National Fish By-Products Working Group' exists which has been acting as a clearing group for solutions in disposing of fish wastes for Irish fish farmers and fish processors. Further a study of fish waste arisings has been conducted for the Marine Institute Ireland and this indicates that Ireland may be a stage further along the process of tackling fisheries waste than Scotland if not any nearer a solution.
Irish fish waste, including both wild fish wastes and mortalities but excluding epizootic events is estimated to be 63,786 mt (Nautilus Consultants (Ireland) Ltd, 2003). Aquaculture waste is 5,543 mt. Most waste arises in the North West region where it is mainly processed by the fish meal plant at Killybegs. This factory ceased taking in salmon waste in 2001. Since that time 50% of this amount has been rendered at Ballynasloe.
Marine Harvest has installed composting facilities near their processing unit in the north and claim to have halved the 92/mt cost of rendering and subsequent incineration. It is not yet clear whether it meets the requirement of the ABP of 1 hour at 70 °C. The feed stock for this unit is process waste and morts. The Irish industry is still recovering from a Radio Telefís Éireann (RTÉ) programme in October 2003 that showed one farmer burying morts in a bog and a farm with morts lying unrecovered under fish cages.
Major epizootic events in Ireland have been less frequent and up until the mass kill at Inver Bay in 2003, very much smaller than in Scotland. The 2003 event illustrates the quandary of the industry both in Scotland and Ireland. It is hard to justify investment in facilities to deal with such occurrences due to their erratic/sporadic nature. However whilst Category 3 plants cannot normally handle material from these events, then dedicated units are necessary.
APROMAR, the Spanish marine fish farmers association, advise that although their industry is small, the ABP Regulations is an increasing constraint to waste disposal, particularly for mort disposal from land-based turbot farmers. No waste management solutions exist and it is suspected that dumping may occur or that morts go for bait and to tuna ranching which has become popular in Spain. They do not have problems with process waste because all sea bass and sea bream go to retail outlets via wholesale markets in the round (ungutted) form. The problem therefore is for the retailers. Renderers are reported to be reluctant to take dead fish in Galicia because of the high water content. Overall it looks as if Spain is further behind with a solution than the UK.
In Denmark there are only three fishmeal companies. Two of these, 999 and Skagen Fiskindustrie are amongst the largest in Europe. As in Norway, no aquaculture by-products go to their factories in Esbjerg, Skagen or Thyboron. Denmark's aquaculture industry is almost entirely based on the production of 40,000 tonnes per annum of rainbow trout in the sea and fresh water. Unlike salmon production, the fish spend less than a year in production so there is a marked seasonality to offal supply. Unlike in the UK the fur animal industry thrives and requires large amounts of feed stuffs. Hence a proportion of trout offal goes to this use or biogas production. It is supplied to the fur farms as fresh, frozen and ensiled. Lumino A/S is Denmark's largest ensilage processor/collector with a wide range of products for pigs and poultry. They face some competition from Norwegian suppliers who have little in the way of a domestic market.
Denmark is an important intermediate processor of Norwegian salmon en route to Germany, France and Spain so the amount of salmonid offal available is greater than the 6,500 tonnes estimated to come from domestic aquaculture. The proportion of the total amount which does not go to fur or ensiling is small and is principally that portion which is unsuitable for either use. It is rendered in the DAKA plant where it forms a very small proportion of their throughput. DAKA are Denmark's principal animal renderer and producer of animal protein meals.
2.7.6 Alaska, USA
Alaska differs from all other countries in that its capture fishery includes the large and commercially important, wild salmon industry. The salmon catching sector is passionately opposed to aquaculture and it is unlikely that fish farms will ever be established there. With no aquaculture, separation of farmed wastes from capture wastes is academic. Use of Alaskan meals would not be banned in aquaculture under the ABP Regulations as they contain only fish caught for fishmeal production or by-products of fish for human consumption which are not of the species being fed. However the familial closeness of the species concerned would almost certainly cause misgivings in Europe. The entire production goes below the 49 th parallel for animal feed production. Because Alaskan products must be shipped by US carriers, exporting other than to the US is uneconomic and so much Alaskan fish oil is burnt as fuel.
Alaska has a huge groundfish catch - over 60% of the food fish catch of the USA. The amount of by-products generated from human food processing operations is in excess of 1.1 million mt. Unfortunately unless the processing plant is close to one of the few fishmeal plants, the logistics are such that it is more economical to grind up the waste, steam out to designated dumping grounds and return the waste to the sea. The concern here is that despite being faced with such a huge resource, no new technologies have taken advantage of it so far.
With a level of salmon production less than the Faeroes and not much greater than that of Ireland, the aquaculture industry in East Canada produces insufficient offal to justify investment in other than small scale and well tried solutions to the problem. Connor Bros - a company with a long history in fishing, fish canning and more recently salmon farming operates a fishmeal plant whose main input is cannery waste. This plant is the primary recipient of aquaculture by-products in the east of Canada. It has also processed ISA mortalities in the past. Ensiling of morts is carried out but the only commercial collection and further processing business has ceased operations. Composting of ensiled, but primarily fresh, waste for conversion to organic fertiliser is conducted near Bethel, New Brunswick. The operation would not appear to meet the requirements of the ABP since one reliable source reports regular odour problems and problems with predatory birds. A small amount is frozen for bait and pet food.
On the west coast of British Colombia reliable figures for the amounts and disposal routes of aquaculture waste are difficult to obtain. All that can be established is that most waste is processed at West Coast Rendering Inc a rendering plant on Vancouver Island. Ensiling of morts is common. However the volumes were insufficient to sustain two boat-based collectors using ensiling barges, both of whom have ceased operations.
Chile produced 506,000 tonnes of salmonids in 2002 and exported 331,000 tonnes as gutted fish and fillets (Salmon Chile - Estadisticas Dec 2003). Assuming low domestic consumption then aquaculture waste was no more than 175,000 tonnes. The Chilean salmon industry is concentrated on Chiloe Island and the region around Puerto Montt. Region XI to the south of Chiloe is the region where future expansion is expected to occur. Chile is the world's second largest fishmeal producer. It is not surprising then that this technology dominates the aquaculture waste scene in Chile. Pacific Star has dedicated fishmeal plants in Castro (Chiloe Island), Puerto Montt and Chacabuco. It processed 40,600 tonnes of waste in 2001 in Castro alone and expected this to rise by 50% in 2002 (Company statement in Eurofish May 2002). As in Scandinavia, no salmon waste goes to the mainstream meal industry. Of note is the remark by a company spokesman that "We don't just pick up waste from the production plants but also dead animals from the farms". Pacific Star run a fleet of 30 trucks collecting within a 150 km radius and claims to deal with 75% of all waste in Chile. Suppliers were paid 5 Chilean Pesos/kg in 2001 (about 4.50/mt)
The meals produced are used for all normal animal feed purposes other than salmon feeds. Shrimp feed in Panama is one such outlet. The oil goes to applications other than salmon feeds. These include feeding other marine species such as yellowtail in Japan. It is sold at a very much smaller discount to conventional fish oil than in Europe, suggesting that European producers should cast their marketing effort further a field.