Biogas of waterhyacinth

Dumki waterhyacinth biogass project (report phase 1 1999 - 2002)

Introduction

Waterhyacinth is a waterborn weed. It is fast grower (2.5 ton/ha.a) and a pest in parts of Africa
and Asia. An extensive review of waterhyacinth is given by Gopal (I 987). In limited amounts
(fresh 5 kg/day per animal) it can be eaten by cows and waterbuffalo. Habits in Bangladesh restrict
its use only to the wet season and incidental browsing by animals that graze close to ponds.
Waterhyacinth has been mentioned in various reports to be suitable for the production of biogas.
The gas yield is 60% of the drymass (Chynoweth and Isaacson 1987). In the tests reported in the
literature the waterhyacinth has always been finely chopped and fermented in a heated and well
mixed digester. In 1999 a project was started to confirm the use waterhyacinth as a feedstock to
produce biogas. The objective of the phase I project was to demonstrate the digestion of waterhyacinth
in a nonheated digester in Bangladesh. Further objectives were the determination of the amount of the
organic material digested and of the necessary volume for a household digester.

Phase I experiment

We restrict our approach to what can be done manually using a non heated digester. Our setup uses a
100 I barrel in a batch mode with a fixed bed. In the first batch the barrel was loaded with fresh
waterhyacinth and compost. The use of compost in a tropical setting can not be recommended as most
of it did not decompose further and it contained large amounts of clay that had later to be removed.
The use of cow dung seems to lead to less work and a faster start-up of the digester. The density of
fresh waterhyacinth in the barrel without excessive Dressing is about 15%. The dry matter content of
wet waterhyacinth is IO%. This leads to a dry matter of only 1. 5 % in the barrel. By drying the
plants for about a week in the dry season the dry matter content could be increased to about 30%. At
the beginning of the rainy season the ponds were cleared of waterhyacinth and large volumes waterhyacinth
were available, that had been drying for about 30 days. In fact a higher loading is possible as the
waterhyacinth decomposes to fine fibers that can fill-in the spaces between the plants that have not
decomposed. We are interested in a high loading with dry matter as this reduces the required volume of
the digester and thus the investment. We have removed the roots in the later loadings as they in the
early ones did not decompose and are normally associated with significant amounts of clay. Removing of
the roots can be done when the waterhyacinths are gathered manually.

We have opted for a long retention time of the waterhyacinth in the barrel. This allows bacteria and
fungi to grow and digest the waterhyacinth. It also reduces the amount of handling of the rest material
and gives a higher gas yield, The barrel has been filled according to table 11. It seems that a loading
of about 8 kg per year per 1001 is possible. This would yield 5 kg per year of biogas (50 kg/m3 digester).
The temperature of the barrel we use follows the variation of the seasons. This means digestion is faster
in the hot season and slower in the cold season. A large digester that produces sufficient gas for a family
to cook on will operate on the yearly average temperature and have a more even gasproduction. In 2002 the
experiment was terminated. There remained about 401 wet digested biomass in the barrel containing 4 kg
organic material. Part of this would digest further. A fine organic fraction and some dissolved organic
material remained in the liquid phase. The amount of this organic material was not measured. . Some material
could have composted as not all material was submerged in water during the experiment. It is estimated that
the digestion amounted to 70 % of the organic material. This is 10 % higher than with the 3 7 degree short
retention time tests.

Economic considerations

Various options for cooking in the Indian subcontinent have been compared by Ravindranath (I 997)
(Table 111). Cost for the different fuels in Dumki(Bangladesh) are given in table IV. The local costs
have been converted to USD. For a family of six we assume that 20 personmeals per day are cooked. De cost
for fossil fuel and for wood is the same (USD 0.2 per day). We have as yet only an indication for the costs
of biogas from waterhyacinth. We have looked at the costs of gathering and at the investment costs for the
digester. In an experiment three boys collected waterhyacinth and took the roots off. In one hour they
delivered 40 kg of waterhyacinth. The cost for 20 kg/day (2 kg/day dry mass) is then approximately
USD 0.2 per day.

An digester of 15 m3 is needed. This will cost for a concrete construction about 500 USD ( based on concrete
costs of 10 USD for a pit toilet of 0.3 m3). We use capital cost of 15% (interest and repayment), based on
the 8% interest that Grameen bank charges for house loans. This gives a capital cost of 75 USD a year or
0.2 USD per day. Total costs would then b 0.4 USD per day.

This simple cost estimate indicates that the cost for biogas is twice as expensive as wood, kerosene or
propane. We have not included a number of smaller items e.g. the time needed to put the waterhyacinth in
the digester, rent of the boat and additional cost in the construction of the digester. Maybe benefit may
be added due to the potassium (0. 0% to .5 % dry mass waterhyacinth) and phosphor (O.6% to 4.0% dry mass
waterhyacinth) in the water and residue. It seems to me possible by further development to improve the
collection efficiency by a factor of two (e.g. by not taking away the roots) and to reduce the capital
costs also by a factor of two.

Concept of a digester

The concept we would like to test is a digester build up of concrete rings like they are used for rural
toilets but bigger so as to give a height to diameter ratio of about one. The idea is that the digester
can be bottom fed with dried waterhyacinth. This will slowly mix by its buoyancy. The gasproduction will
also add to the mixing process. Our experience shows that after a while at the bottom of the barrel there
is only water with fine solids, but no fibrous material and it should be easy to push waterhyacinth in
this water. The use of dried waterhyacinth means less weight has to be transported and the residue (filtered
water) has a higher content of fertilizer. During normal operation only water with fine solids will be drawn
of Only once every three or so years the whole digester has to be cleared out. I expect that by continuously
loading of the digester less volume will be required as the soft material in the stems will decompose in the
first week and the volume of the fresh waterhyacinth will be reduced significantly.

The introduction of biogas for cooking purposes in Bangladesh may be difficult, as it is a foreign idea.
One option, involving higher capital costs, is to compress the biogas, remove the carbondioxide and put it
in I 001 cylinders. Then its use becomes very similar to that of propane.

Conclusions

Waterhyacinth decomposes anaerobically. The process, giving of biogas, is sufficiently fast as to warrant
further development.

Literature

Chynoweth, D. And Isaacson, R. (I 987) Elsevier Amsterdam
Gopal, B. (I 9 8 7) Waterhyacinth Elsevier Amsterdam
P,avindranath, N.H. "Energy options for cooking in India", "Energy Policy" 25, no I pp63-75 1997

Table I Drying of waterhyacinth

The amount of dry mass has been determined by weighing a sample of the waterhyacinth before and after
drying for at least 12 h at 100 oc.

Table II Waterhyacinth loading of 100 I barrel

Table III Cooking options in the Indian subcontinent

Table IV costs of different fuels in Dumki

Table V cost for a family

Appendix A Chronology waterhyacinth biogasproject Dumki 1999

back