Get our free book (in Spanish or English) on rainwater now - To Catch the Rain.
Biogas from Coffee Wastes
Ken C Calvert http://www.coffee.20m.com/
Jan C von Enden http://www.venden.de/
The Renertech process for making biogas from coffee waste waters was originally developed in Papua New Guinea, whilst the Author was Principal Research Scientist for Coffee Processing at the PNG CIC Coffee Research Institute. It is now under further development in Vietnam, at the Khe Sanh Coffee Factory in the Quang Tri Province. The project is funded by German Aid. Apart from coffee, the process is good for handling large volumes of waste water containing medium levels of organic matter which can be fermented down to organic acids and processed at ambient temperatures
In a matter of hours rather than days.
The Coffee production industry is, in general, not noted for its care of the environment.
Our key word is a generated one, ‘CLEMEG’: Clean, Lean, Mean & Green.
The underlying philosophy for developing this integrated system was to use only naturally available substances, and use them in the most economical way to allow for its use in remote underdeveloped areas with a minimum of infrastructure and supply...
The basis of the process was the discovery that in starting from the traditional two stage UASB process a new second stage could be inserted, whereby a surplus of ground raw limestone or marble chips could be used to automatically buffer a SCVFA solution, (short chain volatile fatty acid), largely acetate, at pH 6.1. At the same time, over a period of more than five years, a mix of psychrophylic methanogenic bacteria were isolated from coffee soils, to which had been added the gut contents of many cold blooded species of fish, reptiles and insects. As a result of prolonged enrichment techniques, we now have a septic strain of anaerobes whichwill gas freely on only coffee waste waters at low pH and at ambient temperatures. This anaerobic sludge now constitutes a valuable resource for the global coffeeindustry. It is the intention of the discoverers not to patent this process in any way and to freely disclose any new developments that come to hand, so that the coffeeindustry of the third world may benefit. However, for those companies and institutions who want to short circuit the 5 year development period, supplies of the Renertech Sludge can be made available under a licensing agreement.
The use of limestone chips as an acid neutralizer for a three stage UASB process, also provides the potential for reduction of carbon dioxide in the output gas by taking out half of the CO2 at the fermentation stage. The traditionally understood reaction for production of biogas starts from acetic acid and produces equal amounts of methane and carbon dioxide.
CH3 COOH = CH4 + CO2
However, neutralizing the acid first with raw limestone produces a molecule of carbon dioxide in the first stage, which can be got rid of before the effluent enters the biogas digester.
2CH3COOH + CaCO3 = Ca(CH3COO)2 + CO2 + H2O
Then, reacting only the acetate ions, produces only one molecule of free carbon dioxide as against two of methane. This makes for a raw biogas with a much higher energy level. While this is of little gain to anyone aiming at a stripped natural gas, for low cost ‘Village’ or ‘Institutional’ level operations with straight biogas ‘per se’, this is a considerable advantage.
Ca(CH3COO)2 + H2O = 2CH4 + CO2 + CaCO3
However, as the solubility product’s of other calcium salts, principally phosphates and a calcium/magnesium complex called ‘Struvite’, are much less by several orders of magnitude, the above carbonate reaction never gets enough calcium ions left over to allow it to go to completion. What can be said however, is that the readiness of the high levels of calcium ions in the reaction to precipitate in one form or another, does encourage the formation of relatively heavy granules which allow for a much faster rate of effluent flow through the digester without losing active material. This would encourage us to promote the EGSB process over the UASB, but formal trials have not yet been carried out. The practical outcome nevertheless is that the biogas coming off a ‘Renertech Process’ digester, is much richer in methane than a typical UASB reaction. This has allowed it to be fed directly into a diesel dual fuel engine without the necessity to strip the remaining carbon dioxide from it first. The wet gas is simply passed through a bed of metallic iron, to wit a drum full of bashed up rusty tin cans, to strip out the sulphides and reduce the moisture levels. This iron sulphide process is completely self regenerating and very simple. Once one digester is working in a new area, the high output of granular sludge seed material, due to the struvite precipitation process, makes start up of further digesters only a matter of days in stead of weeks or months.
To achieve a 6-8 hour process turn around of waste water within the coffee industry, it is necessary to concentrate the processing waste waters by intensive recycling. Every six to eight hours a new tank or silo should be used to store the pulpage,(freshly pulped beans), and a fresh batch of water is used to restart the process. For the next six hours that water, plus all the makeup water required, is drained out through the bottom of the tank of pulpage and recycled back to the machines and the levels of sugars and enzymes allowed to build up to the point where the water is heavily coloured and almost soupy.
This means that all of the pulpage, particularly that at the bottom of the tank, has received the same dose of concentrated pectolytic enzymes at a temperature several degrees above ambient, caused by the recycling water system. What ever hour before midnight the silo has been turned off, fermentation will be complete and it will be ready to be fully washed out the following morning. For those factories that are running a ‘South American’ semi washed type process, there will be a very significant lift in quality because this process gives a ‘fully washed’ output.
The overall process includes a full environmentally friendly clean up of wet factory waste waters. The Khe Sanh factory started off with a very high water usage system using a pair of Pinhalense DC3/6 pulpers and demucilators in a semi washed process.
This plant was converted into a fully washed process by recycling the factory water supply, pumping the demucilated coffee up into a stainless steel silo and allowing the washing and pulping water to drain down through the silo for up to six hours of pulping.
The water from the demucilators was discharged directly to the first stage fermenter or acid pond. Every six hours the pulping water was changed and the coffee pumped to an alternate silo. The first silo was back flushed with water which was used to kick off the next shift of pulping, the coffee was then left to soak under clean water. The discharged recycle water, from the previous six hours, was then also
sent down to the acid pond, a long narrow concrete tank of approx 200 cubic metres, sized to hold around one days throughput of heavily recycled wash water and mucilage from pulping more than 100 tonnes of cherry. When each silo full of coffee was given a further wash the following morning, 8-10 hours after pulping, only very clean fully fermented and washed wet parchment was discharged.
As well as the build up of sugars and pectolytic enzymes in the recycling wash water, there was also as significant rise in temperature.
By the time that the dirty water has flowed down the full length of the fermentation or acid pond , around 15-20 hours, the pH has dropped to 3.8, and all the mucilage has come out of solution and floats as a thick orange scum which is allowed to build up on the surface for several days and turn into a thick black crust which can be raked off periodically and deposited with the screened pulp solids for composting. At the far end of the acid pond there will be a clear middle layer of yellow acid water under the mucilage and over the settled solids. This is then pumped on to the next stage of neutralisation. The rate of ‘acetification’ or fermentation to acid can be speeded up considerably by bleeding off a small percentage of this acid water and mixing it back into the intake of the acid pond to create a ‘feedback’ process..
Use was made of an old 25,000 litre steel tank which was three quarters filled with screened 2-5mm limestone chips. Acid wash water is pumped into the bottom of the tank through a manifold and up through about 1.5 to 2metres depth of chips, with a residence time of 1-2 hours. Once again the surface is covered with a foam of CO2 generated solids, mucilage and a fine black material which is considered to be condensed tannins and polyphenolic materials which have proved in the past to seriously restrict the efficiency of the biogas sludge if they were not removed. Once again, the clear solution from over the limestone and under the foam layer , now at a pH of more than six can be drawn off and used for the next stage and the floating layer periodically raked off and transferred to the pulp solids for compost. It is import to have available facilities to flush this tank and stir up the limestone bed sufficiently to strip off the biological film from the chips which will slowly choke off the flow rate over a period of 2-3 weeks. The use of a wide diameter tank with an open top would enable the froth and polyphenolic scum to be removed more easily by raking off the floating material, just like the acid pond.
At Khe Sanh, the major part of the neutralized wash water is presently discharged into a constructed wet land made in three sections, which will be described later. Because of financial constraints, only a 5000 litre pilot scale UASB digester is working at present. This consists of a 3.5 metre high stainless steel tank. Over the inlet manifold in the bottom of the tank is a layer of more limestone chips about 350mms deep. Above that is the sludge layer which can be up to 1.5 metres deep when inactive, but fluffs up and granulates to make a 2 metres plus deep bed of activated sludge. This sludge will settle and remain quiescent for up to 12 months at a time. However at the beginning of the next season it will reactivate in about a week. The top portion of the tank contains the gas/solids/liquids separator about 500mms under the surface of the discharge water. It is believed that the EGSB process, using a taller digester would be a logical progression over the present system, but this has not been tried yet. It is planned to build a new larger digester of ‘ferro-cemento’ materials for the coming season, which will incorporate these improvements. Trials are presently under way to give some practical numbers as to gas production against tonnes of cherry, rather than the theoretical and difficult to determine kgs. of dried volatile solids etc. With the great variation in effluent strengths, Cherry is the only real measure of inputs into the system for practical evaluation.
The discharge effluent from the digester passes through a small settlement tank, mainly to collect and recycle escaping sludge, and then flows by gravity to the afore mentioned wet lands. The first pond , because of the present heavy discharge from the neutralizing tank as compared to the biogas digester, still carries a lot of BOD and is not a good environment at present for growing anything. Only a few reeds and rushes survive.
The second pond has been planted out with local varieties of hollow stemmed reeds and rushes. These plants actively pump enough oxygen down to their roots to allow them to survive in a totally anaerobic environment, and they are good reducers of both BOD and COD. After the biogas digester, they constitute the second line of biological filtering. In colder climates much greater use would have to be made of the reeds and rushes because our third stage is relative only to tropical climates. The tertiary filter pond is much deeper, 1.5 metres, and is filled with floating water hyacinth which, if the pond is big enough, should take out the majority of the fertilizer salts, the nitrates, potassium, condensed tannins etc, and any remaining phosphates. At our present stage of development however, there is still too much BOD coming from the acetates being discharged from the neutralising tank, not to mention the unreacted acid from the acid pond, to allow the water hyacinth to thrive. They are only just surviving. When they do require thinning out, there are several options available to utilize the excess material, of which the easiest is to chop them up and add them to the composting mix! More biogas, animal feedstuffs and SCP are also possibilities The compost is used as fertilizer to return to the coffee.
Although this process has been developed specifically for the coffee industry, it has also kept abreast of developments in the olive processing and the red wine industry. It is believed that it could be adapted to any fruit or agricultural products industry which has problems with COD anthocyanins and high levels of fruit sugars in large volumes of cold waste processing waters.
Along with this attempt at setting up a fully sustainable coffee processing industry, the further use of Vetiver grass over a period of several years, to create a terraced coffee growing system, with no mechanical earthmoving required, could lead onto an environmentally friendly yet fully mechanized harvesting system which can convert much of the drudgery of excessive hand labour into much more pleasant working conditions and a renewed image of coffee as a progressive forward looking industry.