Part-Load Economy of Diesel-Electric Generators

Abstract

Diesel-electric generators used to produce power in rural Alaska are often found to be inefficient and suffer from premature mechanical failures. Such failures are commonly caused by hydrocarbon build-up in the engine resulting from long-term operation under light-load conditions. There are several feasible approaches to this problem which use proven technology. The most technologically direct approach is to properly size systems. Another involves the optimum control of engine oil, coolant and intake air temperature with thermostatically-controlled electric heaters. Economic analysis shows that this approach could save as much as $13,000 per year in the cost of electricity for a 100 kw diesel generator operating at 25% load. However, further research is needed to establish that the mechanical problems associated with part-load operation are actually abated with proper control of operating temperatures. Practical experience implies that this should be the case. Acoustically tuned low restriction intake and exhaust systems are also an attractive approach because they provide a definite increase in efficiency under all operating conditions. However, these units must be developed for a specific engine and operating speed range. They are not presently commercially available, but could be developed in a continuing research effort. Parallel operation of small diesel-electric generators was suggested by many vendors and operators as a method of improving part-load performance. Though it has the benefit of redundant reliability, the economic analysis does not show a clear advantage because of higher electrical costs near full-load conditions. At very low loads, single small units may also suffer from the same mechanical problems as the large units. The other methods of improving part-load performance which include the use of improved injectors and microprocessor-controlled injection pumps are not presently feasible. However, the state of diesel engine technology is changing so rapidly that these items could become feasible in less than two years. These developments should be monitored closely.