Pictures of the Future      Spring 2008

Optimized Planning

Siemens provides solutions for optimal operational planning at large power plants and for decentralized energy generation. Software helps plant operators make the right decisions and plan daily operations.

Every day, power plant operators have to make decisions based on a range of variables. Say, for example, the weather forecast is predicting blustery conditions, then the chances are that lots of wind power will be fed into the grid, capping demand for energy produced from fossil fuels. On the other hand, will the wind farms be able to cover peak demand over lunchtime, or is it better to rev up a gas-fired plant? Alternatively, it may even be cheaper simply to purchase the extra energy on the electricity market, although prices fluctuate continuously there…

Other considerations that must be weighed include commitments to purchase power from certain types of energy sources, the general state of the grid, and the trade in CO₂ emission credits. The operator’s fundamental objective here is to balance output against demand. In the past, this was largely a matter of experience, though there were also fewer factors to be taken into account. That was back in the days before market liberalization, when it was possible to predict the loads to be generated with a fair degree of accuracy, and when there was neither price fluctuation nor power from renewable sources of energy, where capacity is difficult to forecast.

Today’s power plant operators have to factor in many more variables—but they also have IT support. "Siemens has been supplying IT applications to help large power plant operators with load management since the beginning of the 1990s," explains Erich Fuchs, head of the business segment Decentralized Energy Management at Siemens IT Solutions and Services (SIS) in Vienna, Austria.

One such solution is used for planning periods ranging from one day to a week. Another, called Resource Optimization, was developed in the mid-1990s and is designed to cover periods between one week and 15 months. This helps operators make fundamental decisions regarding the type of fuel to buy and the right maintenance intervals for their power plants. "Whereas short-term planning is based on deterministic weather and load forecasts, long-term planning also includes probability distributions that take uncertainty into account," Fuchs explains. The resulting data is collated online and processed in the power plant operator’s control center.

In an ideal situation, both solutions will be in operation. In the short term, for example, it may be more economical to start up a power plant and burn fuel. On the other hand, the long-term planner might report a limited amount of fuel available for the coming year and advise buying power. If run in combination, the two help to cut costs. "Putting a precise figure on the cost-cutting potential is difficult because of the very different types of power plants in operation," says Dr. Thomas Werner, product manager at Siemens Power Distribution Division in Nuremberg. Yet given the vast amounts of fuel involved, even a few tenths of a percent can deliver big savings.

Virtual Power Plant. In the wake of deregulation and the advent of new technology, decentralized power generation is becoming more and more important for the European electricity market, where there is now an increasing focus on renewable sources of energy as well as cogeneration of heat and power. With renewable energy, however, it is difficult to make precise predictions regarding generating capacity. After all, who can say with absolute certainty how long the sun will shine or how powerfully the wind is going to blow over the next few days? One way of cushioning such fluctuations is to combine several small generators into a virtual power plant. In addition to increasing market clout, this also makes it possible to achieve more accurate predictions and more flexible control of output.

To fully exploit such advantages, however, a virtual power plant requires an intelligent energy management system. "That’s where the Decentralized Energy Management System comes in," says Werner. DEMS is used to enhance an area’s power supply on the basis of predefined economic, environmental, and energy-related considerations. If a virtual power plant is made up of a number of wind farms, the operators are supplied with weather forecasts for wind strength and direction at their particular location. DEMS then draws up an operating plan on the basis of this data and other parameters used to predict the probable generating requirements for a specific region. The system takes into account all the various options for controlling demand and then suggests different strategies, such as powering plants up or down, enhancing buffer capacity, and many other options. The operator can model a variety of scenarios and has a clear onscreen view of all the parameters involved. These include forecast loads, supply and purchase commitments, and the operating schedules of all plants under consideration. "This way, the operator can assemble an optimal operational strategy for all of his plants," explains Werner.

DEMS initially entered service in 2003 at SAPPI Austria Produktions-GmbH & Co. KG, an Austrian paper and pulp manufacturer that uses its own small virtual power plant. An intelligent software package was urgently needed due to the company’s obligation to conform to supply contracts for electricity and gas, its commitment to purchase certain amounts of coal and lignite, and its need to use biomass produced during its own manufacturing. DEMS calculates individual load forecasts on the basis of a production schedule as well as the manufacturer’s own forecast production of power for a maximum of seven consecutive days, in 15-minute intervals. This has made operating schedules much more exact.

Green Power. "There has been a significant increase in the number of customers inquiring about our Decentralized Energy Management System," Werner reports. However, the use of such a solution and the business model behind it are dependent on the national energy policy of the country in question. In Germany, for example, the future is promising. The country’s Renewable Energy Act of 2000 is intended to promote use of power plants that run on renewable sources of energy. As part of a range of measures designed to reduce Germany’s reliance on fossil fuels and imported energy, it also serves the purpose of protecting the atmosphere. The German government plans to increase the proportion of electricity generated from renewable sources to at least 20 % by 2020. By way of comparison, this figure was at around 12 % in 2006. In the same year, total domestic revenues generated with renewable energy—whether from biomass, the sun, or wind—amounted to approximately €22.9 billion.

"We’re not going to see a global solution for energy production," Werner predicts, "but decentralized power generation is set to play a major role alongside the large power plants." And one thing’s for sure. "Such plants will need an intelligent energy management system." And that will ensure optimal use of all the energy sources available.
                                                                                                                    Gitta Rohling