The design and implementation of a demonstration supplementary control system

Abstract

The overall goal of the Chestnut Ridge Supplemental Control System (SCS) demonstration project was to demonstrate how an existing monitoring network, existing air quality models, and existing meteorological forecasting methods could be combined with a new control strategy to integrate SCS into electric power system operation. This final report covers the period February 1, 1974 to May 31, 1976.

A complete SCS for four power plants in the Chestnut Ridge region of Pennsylvania was implemented. The design is described in Section 2. The demonstration period, discussed in Section 3, showed that it is definitely possible to integrate a sophisticated SCS into electric power systems operation. The basic methods used in this project are felt to be directly extendable to other situations.

The only new technology originally envisioned for this project was the control strategy which decides the power system's response to predicted or potential violations. One of the key problems was the need for the control strategy to ensure that standards are not violated in spite of the presence of uncertainties in predicted ambient concentration levels. As discussed in Section 2.6, the implemented control strategy accounted explicitly for the uncertainties.

The only new technology originally envisioned for this project was the control strategy which decides the power system's response to predicted or potential violations. One of the key problems was the need for the control strategy to ensure that standards are not violated in spite of the presence of uncertainties in predicted ambient concentration levels. As discussed in Section 2.6, the implemented control strategy accounted explicitly for the uncertainties.The point source air quality model-used during the demonstration period was primarily a state-of-the-art model. However, as discussed in Section 2.5, a relatively new innovation involving downwash modeling was critical to the success of the demonstration.

During the course of the project, a large data base of SO concentrations, meteorological measurements, weather forecasts, and power systim data was established and stored in a manner which was easy to access and manipulate. Studies were done using these data, both before and after the actual demonstration. Some of the methodologies used and developed are applicable to a variety of problems including many non-SCS types. The results of these studies will now be summarized.

State-of-the-art air quality modeling was not as satisfactory as initially hoped in coping with the rough terrain in the Chestnut Ridge area. Research on improving point source air quality modeling for rough terrain was successfully undertaken using the data base after the demonstration period was over. The results are discussed in Section 5 with details provided in Appendix E.

The Chestnut Ridge area was discovered to have an unexpectedly high background SO level. The data base enabled this background problem to be addressed in %he four ways summarized in Section 4: mean concentration analysis, peak concentration analysis, EPA Larson method, and stochastic modeling. All four approaches are felt to be applicable in other situations where it is desired to understand the true nature of a background concentration. The stochastic modeling appears to be a new methodology with particularly great potential. Details of these four methods are given in Appendixes A, B, C, and D.

Uncertainty arising from air quality modeling errors, weather forecasting errors, fuel sulfur contents, power system economics, and plant availability plays a central role in SCS analysis, design, and implementation. A systematic analysis methodology was applied to the data base to explore how these various uncertainties propagate through the overall SCS and affect its operation. This work is discussed in Section 6.

The control strategy minimizes cost subject to the constraint that ambient standards are not violated. Because of the uncertainties, the control strategy operates in a conservative fashion, that is, it often takes control actions that would not be required if the uncertainties did not exist. The control strategy was applied to the data base to determine how the overall economics behave and how they are affected by the presence of uncertainty. These results are discussed in Section 7.

During the course of the project, opinions were developed on the potentia l future role of SCS. We feel that SCS provides a viable tool for dealing with the energy, economic, environmental crisis. These opinions are discussed in more detail in Section 8.