This figure shows non-dimensional NOx concentration with reference to the NOx concentration at supply air temperature of 42℃. The abscissa axis represents the temperature of 40℃ to 60℃. The absolute humidity of the air is 10g/kg. As supply air temperature increases, NOx13 at high load levels also increases but at a higher rate than NOx13 at low load levels. Under a high or medium engine load, NOx concentration increases 10% when supply air temperature rises 10 ℃. Under a low load, increase in NOx concentration is not obvious in connection with increase in the temperature, compared with under a high or medium load.
5. DEGREE OF CONTRIBUTION OF FLUCTUATION FACTORS ON CGE1 AND SCE
5.1 Fuel Oil Properties
Based on the results of analysis discussed in section 3.5, Fig.9, 10 and 11 present the fluctuation of NOx correction coefficient KHDIES with respect to fluctuation in fuel oil density (F1), cetane index (F2) and CCAI (F3). To compare the degrees of contribution between the fluctuation factors, the analysis was conducted supposing that no other factors have effect. The abscissa axis represents the range of fluctuation seen in the data sampled over one year for the individual factors. For the calculated cetane index, the findings were compared with the results of a test that used different types of fuel oil, conducted on SCE [3]. In case of CGE1, NOx-concentration correction was 5% maximum as far as each factor fluctuates in the range observed in the yearlong study. Based on the results of the above test using different types of fuel oil, when the calculated cetane index decreases 5 from the reference value 43.8(the average of sample data collected from cogeneration engines), NOx concentration increases approximately 5% under a high load of 100%. However, when the engine load is in a low range (25%), NOx concentration increased about 10%, showing that the smaller the engine load, the more effect the fuel oil properties may have. In case of CGE1, the average engine load is 87.8%; therefore, the contribution of fuel oil properties to NOx is relatively low.
5.2 Absolute Humidity, Ambient Temperature and Intercooled-Air Temperature
Fig.12 and 13 show how the NOx-concentration correction coefficient KHDIES varies according to absolute humidity, ambient temperature and intercooled(or supply) air temperature. Regarding absolute humidity, the required correction is almost the same between in the high load range (100%) on SCE and in the average load range (87.8%) on CGE1. The required correction is between -15% and 30%. Of the correction factors examined in this analysis, absolute humidity has the greatest impact on NOx. The KHDIES in the 25% engine load range on SCE was close to that obtained by the formula of NOx Technical Code. Compared with that in the high-load range, the KHDIES was about half. Regarding intercooled-air temperature, the required correction in the low-load range on SCE was almost the same as that obtained by the formula of NOx Technical Code. In the high-load range, however, the required correction was around a significant 15% maximum for SCE and CGE1. The correction required for ambient temperature was almost the same on CGE1 as when it was calculated by the NOx Technical Code formula. Compared with intercooled-air temperature in the high-load range, the impact on NOx was small. When correction required for NOx regarding ambient temperature is negative, that required for NOx regarding intercooled-air temperature is positive, and vice versa.
Fig.13 Room and Intercooled-air Temp. and KHDIES
6. CONCLUSION
This analysis presents the following conclusions.
(1) The fluctuation of NOx concentration from two cogeneration engines was 400 ppm to 500 ppm in one year.
(2) Of the NOx-concentration correction factors discussed in this study, absolute humidity has the greatest impact. Compared with the value calculated by the formula of NOx Technical Code in the high-load range, the required correction is about double.
(3) When correction required regarding ambient temperature for NOx-concentration fluctuation is negative, that required regarding intercooled-air temperature is positive, and vice versa. In the low-load range, the absolute value of the correction required on intercooled-air temperature is about triple that for ambient temperature.
