Skip to content

Long Term Optimization

EnCompass has numerous options for simplifying the detailed commitment and dispatch problem as needed in order to perform multi-year simulations with reasonable runtime.

This may include aggregating hours within a day and using typical days-of-the-week or typical on-peak/off-peak days for each month, but still maintaining the chronology needed for enforcing ramp rates, storage limits, and commitment requirements.

EnCompass allows runs to be segmented with overlapping periods, so that a one-year simulation is a succession of 7-day simulations with an extra day added on to reduce end effects, and initial conditions of subsequent weeks set accordingly.

The level of precision of the unit commitment logic may also be adjusted to produce faster simulations, but still account for commitment costs and ancillary service requirements.

Time Series

All costs, volumes, and constraints in EnCompass are input using a Time Series, which can be just a single entry or multiple future entries that repeat or escalate throughout time.

Certain inputs such as Demand and Energy Prices that have daily shapes defined may be escalated annually (maintaining the proper days of the week) while being scaled up or down to match inputs for annual peak and energy (for Demand) or monthly on-peak and off-peak (for Energy Prices).

Planned maintenance for resources may be entered with historical or planned dates that repeat in the future every 12 or 18 months, for example. Each potential outage may also have an allowable shift, which allows EnCompass to determine the optimal timing of planned outages within a region.

Unplanned outages can then be randomly determined for each resource based on input forced outage rates and lengths. The structured scenarios within EnCompass make it easy to use the same outage schedule from a long-term parent scenario to multiple child scenarios.

Anchor Power Solutions | EnCompass Software | Power Planning Software | Optimization

EnCompass uses Time Series and outage scheduling together with parallel processing and risk reporting to quickly generate a long-term forecast from near-term conditions.