RER Renewable Energy Research  français
   Overview       Expertise       Publications       Resources        Photos    
    All       By Topic       Articles       Reports       Non-technical       Presentations    

Analysis of First Hybrid Test Bench Experimental Dispatch Strategy

Michael M.D. Ross
RER Renewable Energy Research

Full Text Report
Link to CETC-Varennes


Research conducted on behalf of the Photovoltaics and Hybrid Systems Program at the CETC-Varennes (Natural Resources Canada) with partial funding from the Panel on Energy Research and Development (PERD).


Ross, Michael M. D. Analysis of First Hybrid Test Bench Experimental Dispatch Strategy. Report to CETC-Varennes (Natural Resources Canada). Montreal, Qc: RER Renewable Energy Research, 2005.


The CETC-Varennes hybrid system test bench was used to evaluate a relatively novel genset dispatch strategy for PV hybrid systems. The dispatch strategy involved starting the genset on the basis of an estimate of the state-of-charge; running the genset for a variable amount of time, such that less charge was returned to the battery in the early morning, in the expectation that battery capacity will be necessary for storing solar energy during the day; and achieving regular full charging of the battery over a mult-day period using a combination of photovoltaics and genset charging, thus avoiding long genset-powered absorb charges.

Unlike in previous tests, where peak loads depressed the battery voltage and caused genset starts at relatively high states-of-charge, in this test the genset start times were more closely correlated with the state of the battery than the load. Of particular interest, the voltage-SOC relation was adjusted based not on the instantaneous load current, but rather on the basis of the recent peak load current. It was found that when only the instantaneous peak current was considered, the genset tended to be dispatched immediately following a peak in the load current. This occurred due to the slow recovery of the battery voltage following a step decline in the load current.

Adjusting the genset run time based on the time of day, in order to keep battery capacity available for photovoltaic output on the upcoming day, is not a particularly rich vein of potential performance improvements. During high solar fraction periods, the genset will rarely run in any case; during low solar fraction periods, the battery will have tendency to discharge, and keeping battery capacity available for surplus solar energy will rarely be a problem. Periods of intermediate solar fraction will tend to be transitory.

A simple alternative to adjusting the genset run time based on the time of day is the adjustment of the genset start criterion on the basis of the solar fraction. A low SOC threshold should be used during periods of high solar fraction, such as summertime, and a high SOC threshold should be used during periods of low solar fraction. This will raise the average state-of-charge when the solar fraction is low but will not lead to significant waste of solar energy.

The use of PV and genset to achieve full charging over a period of multiple days yielded particularly interesting results. Past tests have shown that partial state-of-charge cycling causes a cycle-on-cycle deterioration in the apparent capacity of the battery. In this test, this same effect appears, but in reverse: over a multi-day period, the apparent capacity of the battery is increased by repeated cycles of running the genset in the morning, letting the PV charge during the day, and then operating the load through the night. In short, we observe a particular form of partial state-of-charge cycling inducing an improvement in the battery state, to the point that full charging is achieved. This presents an attractive avenue for reducing the time spent by the genset in absorb charging, which is inefficient and causes wear.