Experience with Installing and Using Optimization Software

Mighty River Power uses optimization software to improve overall performance of its nine hydro facilities. Based on results and lessons learned from initial operations, the utility plans to investigate the expansion of functionality within the system to dispatch units based on river flow and generation requirements.

By Andrew B. Spackman and Richard B. Allen

This article has been evaluated and edited in accordance with reviews conducted by two or more professionals who have relevant expertise. These peer reviewers judge manuscripts for technical accuracy, usefulness, and overall importance within the hydroelectric industry.

Mighty River Power owns and operates nine hydro plants with a total of 39 turbine-generator units on the Waikato River in New Zealand. The stations, with a total capacity of 1,040 MW, operate in a cascade; water from the upstream generating plants flows downward to the other plants.

For some time, Mighty River recognized that, by running its hydro units at their most efficient point for any station load, additional value could be extracted. However, multiple demands on the plant operators (referred to as trading group dispatchers) and the rapidly changing nature of Mighty River’s generation profile made this difficult to achieve with its manually driven remote-control system. Therefore, Mighty River decided to incorporate an optimization software system into the existing control system.

Choosing optimization software

For this software solution, the utility wanted a proven system that could be serviced and supported in New Zealand. The software needed to provide real-time solutions and interface seamlessly with all of the utility’s existing production systems and databases.

After reviewing the available options, Mighty River Power chose the RT Vista software from HATCH Renewable Power — Power and Water Optimization (SGI), a decision support system (DSS) provider for the hydropower industry.

The optimization software maximizes the efficiency of generation units at a station level. At the same time, it fits into the constraints of the New Zealand energy market, the existing supervisory control and data acquisition (SCADA) system, and established unit dispatch processes. The objective of the DSS software is to provide seamless integration from the point of entering a station dispatch quantity into the control system to the automatic starting, stopping, and loading of the units within the station to achieve the required dispatch quantity at maximum efficiency.

Flexibility is a key benefit of the software. It enables dispatchers to solve problems by performing analysis on a station-by-station basis, or on an entire system. Dispatcher controls provide two modes of automation: advisory and control.

General requirements

Because the outcome of this optimization project involved close interaction of staff as well as close interfacing with existing software systems, its success depended on both personnel and technical requirements.

Personnel requirements

From a dispatcher’s perspective, a new tool had to be both easy and fast to use. There were already many demands on the dispatchers’ time; therefore, optimization needed to be accomplished with as little effort and time delay as possible. RT Vista needed to be nonintrusive to the established methods used to run the plant. The only way to achieve this was to integrate the control and information requirements into the existing system. All commands required to run RT Vista and all information that it provided had to be available within the existing control screens.

One of the goals of optimization was ensuring the dispatchers saw as little change to their existing operating environment as possible. Designers retained the overall look and feel of command menus and information used to manually control the units. From the personnel perspective, this “dispatcher experience” was a key requirement for success.

Technical requirements

The technical requirements of close integration and the software’s response rate were integral in creating a positive user experience for the dispatcher. The information necessary to determine loading of generation units for optimal efficiency was stored in various separate databases. Dispatchers needed the ability to access the information automatically, because manually managing the data was a cumbersome and time-consuming process.

Using the system

RT Vista works continuously to ensure optimum real-time dispatch of hydro generating units in a cascade. The software’s output passes directly to a plant’s control system, either as an advisory recommendation or in a closed loop to the unit’s programmable logic controller (PLC). The operations results also pass to an archiving system to enable reporting of performance.

The dispatcher receives a system load request from the grid operator and resolves this request into individual station power outputs. If the software is in “advisory” mode, then the optimal unit dispatch is displayed on the control system screen and awaits the operator’s acceptance or manual edit. If the software is in “control” mode, then the optimal unit set points are also sent directly to the machines.

Establishing connectivity between the two major systems, SCADA and RT Vista, involved using several different protocols. Controls were designed so that all commands to the field originated in SCADA rather than RT Vista.

Optimizing unit dispatch

The objective of the real-time unit dispatch process is to provide the dispatcher with unit megawatt set points that lead to optimal unit performance. RT Vista guides the dispatcher through the analysis of a unit commitment/dispatch problem. The guidance is provided in two different modes of analysis: optimal solutions based on a prescribed planning horizon and solutions based on an instantaneous time frame.

This software closely monitors and controls the unit commitment and loading applied to each unit within a station. In the initial stage of development, the main operational objective was station optimization of unit commitments and load for the “current” time. It included a view for optimal unit commitment on a five-minute basis over the next four hours.

Click to Enlarge
New Zealand hydro project owner Mighty River Power installed decision support software into the dispatcher’s SCADA system in an effort to improve operational efficiency of the 39 generating units in its nine hydro plants.

The key element for success in the Mighty River optimization project was keeping the dispatchers involved with the design process while not exposing them to modeling roadblocks until after the module was ready for production testing. Maintaining the dispatchers’ faith and trust in the results from the DSS module was paramount in achieving a successful implementation of the software.

Plans for the software

Mighty River Power plans to expand on the optimization already in place by integrating river optimization over the next four hours into the existing optimization methodology. This would include the allocation of system generation requirements among the available generating resources while obtaining optimal hydraulic positioning in the river.

The algorithm would optimize the river flows to maintain stable, targeted water levels over the next four hours. The unit dispatch solution would address many operational decisions such as optimal unit commitment to meet all dispatch requirements, within the bounds of all operating constraints.

Andrew Spackman is the maintenance systems engineer for the geothermal section of Mighty River Power in New Zealand. He served as the project technical manager for the optimization work described in this article. Richard Allen is the head software development/chief technical officer for HATCH Renewable Power — Power and Water Optimization (SGI) in Canada. He served as the technical leader for the optimization work.

More HRW Current Issue Articles
More HRW Archives Issue Articles

Previous articleIDB loan approved for 80-MW hydro project in Bolivia
Next articlePower Machines to supply hydro units to Sayano-Shushenskaya plant

No posts to display