Quality Control in the interim network for the measurement of ammonia: an investigation of the influence of the temperature inside the measurement stations
Kwaliteitscontrole interim meetnet ammoniak: een onderzoek naar de invloed van de temperatuur in de meetstations
Elzakker BG van , Stuiver J , Uden GJBM van
RIVM Report 723101006
Because of doubts concerning the correction method for temperature sensitivitiy of the instruments used in the interim network for the measurement of ammonia (NH3), this sensitivity was investigated again using two correction methods ("old" and "new") as proposed by the manufacturer. The "old" method is used in the interim network and turned out to be principally wrong because it does not use the correct reference temperature while the "new" method does, namely 20 degrees C. Also, the influence on 1-hour averages, period-averages and diurnal variations when using the "new" method for recalculating the excisting NH3 database, which was determined with the "old" correction method, was investigated. A deeper understanding was gathered about the real temperature range in the measurement cabins. The investigation in a climate chamber showed deviations in the 1-hour averaged concentrations up to -25% at 35 degrees C (26 mug/m3, "old" correction method) and +20% at 35 degrees C (1.3 mug/m3, "new" correction method) respectivily. Particularly, a concentration dependence was found for the temperature sensitivity. If no concentration depending correction is used, deviations of -11% up to +20% at temperatures of 10 and 35 degrees C respectively ("new" correction method) can still occur particularly with low concentrations (ca. 1 mug/m3). Below this value, the absolute deviation becomes small. During wintertime, the temperature in the cabins is 20 +-2 degrees C. In summer, at high ambient temperatures (30 degrees C), the temperature in the cabin can reach 35 degrees C at about 19.00 h. So, the air-conditioning is not dimensioned well. With the "old" correction method, which is now in use, one hour averaged ammonia concentrations are underestimated with about 25% at these high temperatures. Period averages (2 - 4 weeks) are underestimated less, namely about 3%. This is caused by the low frequency of simultaneous occurrence of high temperatures and high concentrations. For the same reason, diurnal variations are only marginally influenced. It is recommended (after additional investigation) to make the temperature correction concentration dependent, to improve the temperature-conditioning in the cabins or to thermostat future versions of the instrument itself.