mobile applications like marine and automotive diesel engines.During this period, from the 60s till today, a significant change in the feedstock for the stationary boilers has occurred. Due to the concern about greenhouse gases, the use of renewable fuel has increased, in order to reduce the use of fossil fuel. Renewable fuel can be, for instance, woody biomass and forest residues, municipal, industrial or agricultural waste.While the previous use of fossil gas, diesel or oil produced a rather particulate free flue gas, the new fuels contain larger amounts of alkali metals. Thus, they generate a large portion of fly ash. A recent study compares the use of SCR to SNCR for reduction of nitrogen oxides [1]. The authors say that if the NO x level for waste incinerators would be lowered to 100 mg/Nm 3 , it will become necessary to use SCR instead of SNCR. For new installations, the catalyst should be placed after the boiler in a high dust position. This means that fly ash will pass through the catalyst bed via the monoliths channels, resulting in a close contact between the catalyst and the fly ash. There are many potentially poisonous compounds in the fly ash. Among them alkali metals are of a major concern [2-5]. They cause deactivation of the catalysts, by blocking the Brønsted acid sites, in the same way as for applications burning forest residues [2,6].We have studied SCR catalysts in various applications since the mid 80ies [7,8]. In one paper, we published the effect of lead on the activity [9]. Earlier results from deactivation studies on diesel engine driven power plants in Sweden [10] show that there is an activation after about 900 h on stream. In that case, the conversion increased from 85 to 92% at 350 °C. In an earlier study we showed [11] that gas phase sulphating by 500 ppm SO 2 in 500 ppm NO, 4% O 2 and 4-5% H 2 O with He as Abstract A commercial vanadia, tungsta on titania SCR catalyst was poisoned in a side stream in a waste incineration plant. The effect of especially alkali metal poisoning was observed resulting in a decreased activity at long times of exposure. The deactivation after 2311 h was 36% while the decrease in surface area was only 7.6%. Thus the major cause for deactivation was a chemical blocking of acidic sites by alkali metals. The activation-deactivation model showed excellent agreement with experimental data. The model suggests that the original adsorption sites, from the preparation of the catalyst, are rapidly deactivated but are replaced by a new population of adsorption sites due to activation of the catalyst surface by sulphur compounds (SO 2 , SO 3 ) in the flue gas.