Waterpower XV Notice

Interesting Vista-related papers to be presented at Waterpower XV this summer…

The following two papers will be presented by Synexus Global® staff as associated clients:

Drive for Efficiency: Impacts of Real-Time Station Optimization Software within Mighty River Power

Mighty River Power owns and operates 1056MW of hydro generation from 39 generators spread across nine hydro plants on the Waikato River system in the North Island of New Zealand. The river begins at Lake Taupo in the middle of the island. The first station (Aratiatia) is 13 kilometers (km) northeast of Lake Taupo, and the last station (Karapiro) is 188 km downstream, 30 km southeast of Hamilton City.

The stations operate in a cascade, which is a network of interconnected river reaches with water from the upstream generating plants flowing downward to the other plants. All generators are remotely controlled via SCADA to PLC based automatic starting and stopping systems.

Mighty River Power bids into the New Zealand electricity market which operates as a 100% real time spot market based loosely on the Nordic market. The market trades in 30 minute time slots with pricing offers made 2 hours in advance. The market network has around 200 network nodes, each with their own nodal energy prices. Ancillary markets are in place for spinning reserve and frequency keeping. It is a very small market by world standards and has a high level of volatility in load and prices.

To make best use of existing investments it was realised that additional value could be extracted from the Hydro plant by ensuring that the generators are operating at their most efficient point for any given station load. The responsibilities and work load already managed by the dispatchers (plant operators) and the rapidly changing nature of the generation profile meant this would be difficult to achieve with any manual process. Following a review of other possibilities, MRP came to a decision to achieve more efficient operations by the use of a software solution. To be successful this solution would have to work seamlessly with the existing SCADA system to minimize the workload on the dispatchers. It would also have to incorporate the many constraints and rules under which the river system and individual units are managed. This formed the basis for investigating possible solutions to drive efficiency gains.

Optimizing Ancillary Services in an Independent System Operator (ISO) environment

Southern California Edison (SCE) is an electric utility in the state and supplies power to over 11 million people within a service area of central, coastal and Southern California. The SCE hydro generation portfolio includes the Big Creek hydro generating facilities on the San Joaquin River, consisting of 9 hydro plants with a total capacity of about 1000 MW, a portion of the Hoover hydro electric project and various dispatchable thermal units.

SCE is a full participant in the California Power Market, which includes energy and ancillary service markets. SCE identified the need for a hydro optimization modeling tool to support both long and short term planning of SCE’s energy limited hydro resources, in particular, to maximize the opportunities on the energy and ancillary service markets, subject to operational and water management constraints. Subsequently, SCE acquired and implemented the commercially available Vista Decision Support System to assist in the generation planning and water management of its’ hydroelectric facility.

One of the project tasks was the enhancement of the short and long term scheduling models with a combined optimization scheme for both the energy and ancillary service markets. The reserve transmission and generation model was extended, where differentiation between the various products, i.e., non spin, spin, regulation up and down was fully integrated. This included a detailed physical representation of the generating units and plants in terms of their various reserve capabilities, as well as the relation between reserve and unit efficiency.

A key aspect in optimizing for ancillary markets is the consideration for the short term uncertainty in the amount of reserve that will ultimately be called upon, as called upon reserve affects water availability and revenues. A probabilistic approach was implemented to deal with this uncertainty, with the overall solution providing the plant generation schedule, water usage, and bid transaction summary.

In addition to the work conducted on the optimization models, considerable attention was placed on data integration to arrive at a seamless process that can be implemented in the operational environment. All key inputs from the hydro operation, such as reservoir water levels, were completely automated, as well as all the required information concerning market price forecast.

This paper will provide a description of the various challenges encountered during the project, the concepts used for solving the combined optimization problem and some typical results that were obtained from this groundbreaking development.