EMS optimized management of a CHP: Impact of performance curve uncertainties

In the last years, reduction of energy costs and CO2 emissions are the main actors in the plant management. For this reason, energy market shew an increasing interest in distributed generation; in particular, high efficiency equipment like CHPs are one of the ways to guarantee compliance with EU directives on environmental impact without worsening performance. In order to boost this trend, a second way to improve generators efficiency is the adoption of energy management systems (EMS): intelligent devices capable to provide an optimized schedule of generators taking forecasts of next days loads as input. When an EMS is installed in a generation plant, it is important to configure the correct generators' performance curves. These curves depend on generator type, size and manufacturer; one aspect that cannot be neglected is the difficulty in extrapolating curves parameters from datasheets or from real data. Often, equipment manufacturers provide only generators rated power or, however, not exhaustive information: thus it is necessary to estimate curves parameters from running devices acquiring data directly from field. This kind of campaign, due to its nature, will for sure introduce some issues related to missing data, measurement troubles and noise affection; the effect of these perturbations will cause uncertainties on performance curve evaluation of the generators. The aim of this paper is to analyze the behavior of an internally developed EMS under CHP performance curve perturbation hypothesis and compare it with the behavior of the same EMS assuming that the curve reflects the real functioning of the CHP. The EMS under analysis will be considered running in a real tri-generation plant in the Italian scenario.