Trials by Sumatra Bioscience have shown that high quality compost can be produced by composting empty oil palm fruit bunches with oil mill effluent in an open windrow system over 25 days. Fifteen tonnes of the final product typically contains 105 kg N, 16 kg P, 168 kg K and 26 kg Mg, which is close to the average nutrient levels applied to oil palm in Sumatra as inorganic fertiliser per year, except for P which is lower in the compost. Thus, compost applied alone clearly has the potential to replace the inorganic fertilisers usually applied to oil palm. Two trials have been carried out to compare the effectiveness of compost and inorganic fertilisers applied to oil palm on a typical volcanic ash soil in North Sumatra and to determine the optimal rate and method of application of the compost.

The first trial tested a factorial combination of three rates of compost, urea, rock phosphate and muriate of potash. The highest rate of compost (10 t ha^-1 yr^-1) applied alone increased the FFB yield from 23.1 up to 26.8 t ha^-1 yr^-1 over a three-year period, which is an increase of 0.37 t ha^–1 yr^-1 FFB per tonne of compost applied each year. A similar yield was achieved with 2 kg urea plus 1 kg rock phosphate (there was no response to K fertiliser), which based on the nutrient content of the two materials, indicates that the N and P in the compost were 66% and 37% more effective than the nutrients in the inorganic fertiliser (and confirms that the P content in the compost is more than adequate). The greater efficiency of compost compared to the inorganic fertilisers in supplying N and P to the oil palms was also confirmed by the higher recovery of these nutrients into the palm fronds. The highest yield of 28.7 t ha^-1 yr^-1 in this trial was achieved with 10 t ha^-1 yr^-1 compost plus 2 kg urea and 2 kg rock phosphate, indicating that if only compost is applied, the highest rate will be needed to achieve the optimal yield.

In the second trial, compost alone was tested at incremental rates of up to 20 t ha^-1 yr^-1, applied as a patch between the avenues and as 1, 2 and 3 m bands down the avenues. The optimal yield over three years was achieved with 15 t ha-1 yr-1 compost, which increased FFB production from 26.9 to 32.6 t ha^-1 yr^-1; which is an increase of 0.38 t ha^-1 yr^-1 FFB per tonne of compost applied per year and this is in very close agreement with the response recorded in the first trial.There were no significant differences due to the different methods of application of the compost. As in the first trial, N and P nutrient contents in the fronds were very significantly increased by the compost, but not K, due to a high K nutrient reserve level in the soil.

The cost of producing 1 t of compost by the windrow method and applying it in the field in Lonsum estates has been calculated as USD 10, so the application of 15 t ha^-1 of compost in the second trial would cost USD 150. Application of inorganic fertilisers at the same nutrient rates as 15 t compost ha^-1 yr^-1 at 65% moisture content is approximately twice as expensive as compost. Taking into account the greater efficiency of compost in supplying nutrients, replacement of inorganic fertilisers with compost would save the costs of inorganic fertilisers almost three times as much. However in the trial area, K and Mg fertilisers were not required, so savings by switching to compost at this particular location would be less.