ECN publication
Roterende Deeltjes Scheider (RDS) voor reiniging van biomassa stookgassen
Meijden, C.M. van der; Neeft, J.P.A.; Ploeg, F.B. van der
Published by: Publication date:
ECN 1-9-2001
ECN report number: Document type:
ECN-C--01-118 ECN publication
Number of pages: Full text:
28 Download PDF  

The building, testing and evaluation of a Rotating Particle Separator (RPS) for the removal of dust, tar and NH3 from biomass producer gases is reported. The RPS was operated at a low temperature at which water from the producer gas condenses, hence it is called a wet RPS.In a technical evaluation, experimental tests were performed with the RDS installed downstream of the ECN CFB gasifier. These tests showed that the filter element of the RDS blocked by tars within hours after start of the tests, causing that the RPS had to be stopped. Cleaning of this filter element by a continuous water spray was not sufficient. Optimisation of the cleaning of the filter element is not strived for as during the project Aarding decided to abandon its RPS activities. However, it was also shown that the RPS can effectively remove dust, tar aerosols and NH3 from biomass producer gases. The removal efficiency is estimated to be sufficiently high for trouble-free use of the producer gas in a gas engine. In a rough economical evaluation, the RDS is shown to be a potential more attractive option for biomass gas cleaning than an Electrostatic Precipitator (ESP) or a catalytic cracker (dolomite, TPS concept). However, the cleaning of the filter element must then be solved at limited costs. A nitrogen purge is a too costly option. If a nitrogen purge for cleaning of the RPS is included in the economical evaluation, then the ESP is the most attractive gas cleaning option amongst ESP, RPS and catalytic cracker. For the catalytic cracker, the investment costs are too high. Finally, on the basis of the results of this study it is argued that existing tar specifications for use of producer gas in gas engines (the regularly mentioned 50 or 100 mg/mn 3) can be regarded as not very useful. An alternative specification is proposed.

Back to List