SINTEF Energy Research of Norway is well-known for developing EOPS, the power system simulator that has been used for expansion planning in Norway for most of the major hydro projects since 1980. It has also been used recently in Vietnam, Nepal and Iceland. Approximately 25 medium-sized power producers in Norway use the EOPS model for operation planning.
EOPS is EFI’s One area Power market Simulator programme (EFI changed its name to Sintef Energy Research following a merger in 1998). The programme is designed for long term scheduling and expansion planning in hydro-dominated areas. Assuming that there are no major transmission constraints, and some degree of homogeneous hydrological characteristics within the area, EOPS will calculate incremental water values in a single reservoir model and, under simulation, use a rule-based drawdown strategy to optimise water allocation between reservoirs.
EOPS uses stochastic dynamic programming to model the uncertainty of load caused by temperature variations and inflow. An operator inputs the local spot price, or defines the load and the cost of curtailments, and a description of the system. Each module is described by:
• A reservoir.
• Storable and non-storable inflow.
• Waterways for overflow, bypass and plant discharge.
• A power plant.
• Constraints on reservoir storage and water flow.
Hundreds of plants in the same or different river basins may be modelled within the same system.
EOPS may perform a multitude of modelling tasks, including: long term operational scheduling, maintenance planning decisions, risk management, reservoir operation and the calculation of market clearing prices in isolated systems.
For the simulation of hydro-thermal systems that cover wider, interconnected areas, perhaps with price variations, Sintef has developed a multi-area power market simulator (EMPS). Many of the power producers in Scandinavia use EMPS for price forecasting and to research the benefits of transmission connections. Unlike EOPS, which operates on a single busbar, this system calculates water values for regional subsystems in large interconnected power systems, and simulates optimal operation for a number of hydrological years.
EMPS aims at the optimal use of hydro resources, in relation to uncertain future inflows, thermal generation, power demand and spot type transactions within or between areas.
In the strategy evaluation part of the model, regional decision tables in the form of incremental water values are computed for each of the areas in the system. In the simulation part of the model, optional operational decisions are evaluated for a number of hydrological years (typically between 30 and 60). Hydro and thermal production are determined for each time step in a market clearing process, based on the water values for each aggregate regional subsystem. Aggregate hydro production within each subsystem is distributed among the available plants, using a rule-based reservoir drawdown model containing a detailed description of all plants and reservoirs.
The basic time step in EMPS is one week, with a horizon of up to three years. Within a week, the load may be represented by a duration curve of up to four steps, say, peak, off-peak, night and weekend. The current capability of the model is 900 hydro modules, divided arbitrarily within up to 25 subsystems. Additionally, hundreds of thermal plants may be modelled.
As well as spot-price forecasting, long term scheduling, maintenance planning and reservoir operation, EMPS can calculate the probability distribution of hydro or thermal production potential, and the utilisation of transmission capacity between subsystems. It may be coupled to a detailed load flow model.This allows the user to analyse the utilisation of individual lines and cables and perform detailed analysis of total and marginal losses: a useful tool for bottleneck and grid tariff analysis.
Both EOPS and EMPS are long term simulators. For short-term hydro operation planning, the company has developed SHOP, a programme based on an optimisation formulation, where any number of complex hydraulic configurations or watercourses can be modelled. A typical study period is up to ten days.
The SHOP software includes all the main components such as reservoirs, hydro units, discharge gates and thermal units. Different network areas with separate load profiles and short-term market descriptions are also modelled.
The power plant and the gates can have any destination downstream and can be given individual time delays. The endpoint reservoir level can be described by water values or specified endpoint values. Hydro units are defined by any number of power production/discharge curves, reflecting different upstream and downstream reservoir levels. Linear interpolation is assumed at intermediate plant heads. The discharge gates can be defined by tables where the release is defined as a function of upstream reservoir level and gate position.
The results taken from the software may include reservoir trajectories, power production, discharge and incremental cost.
External weblinkswww.energy.sintef.no
