As states around the country move forward with utility deregulation, it's more important than ever for fossil-fuel-based power generation facilities to find efficient, cost-effective ways to reduce nitrogen oxide (NOx) emissions - particularly those in ozone non-attainment regions and those located in the 19 states that the U.S. Environmental Protection Agency (EPA) has asked for pollution reduction plans.
What Makes Reducing NOx Emissions So Important?
Under the Clean Air Act, EPA categorizes regions based on their ability to meet federal ground-level ozone (smog) standards. Some regions attain those standards, while others fall into various categories of non-attainment. Ozone - which has been shown to cause a variety of respiratory problems in humans, including reduced lung capacity - forms when NOx combines in the air with volatile organic compounds (VOCs) on hot, sunny, stagnant days. Regions that cannot achieve federal air quality standards are subject to economic sanctions, most often in the form of limits or bans on the growth of heavy industry like power generation.
Neural network technology, a type of artificial computer intelligence, can help power plants achieve the critical balance between NOx reduction and boiler efficiency.
One approach to decreasing ozone pollution in non-attainment areas - which EPA is focusing on with the controversial NOx State Implementation Plan (SIP) Call - is by limiting the amount of NOx, a precursor to ozone, present in the ground-level air. NOx is a by-product of the combustion process - the hotter the flame temperature used in the combustion process, the more NOx the process produces. Industrial areas and densely populated regions with long periods of heavy traffic and automobile gridlock can generate higher levels of NOx than other geographical locations, and are often classified as non-attainment regions for ground-level ozone.
State governments, which bear the burden of meeting EPA's Clean Air Act standards, have continually turned to fossil-fuel-fired power plants and other industries using large boilers as the primary opportunity to reduce NOx emissions and bring their regions into attainment with federal ozone standards. Under the proposed NOx SIP Call, 19 states will face the burden of cutting NOx emissions even further, in order to relieve Eastern states of increased pollution due to ozone transport over state and regional lines from upwind areas. Many states are again looking to fossil-fuel-fired power plants for the most dramatic NOx cuts.
On top of this ongoing environmental challenge, the electricity marketplace is experiencing a revolutionary change. One by one, states around the country are deregulating electric generation utilities, allowing end-users, including businesses and homeowners, to choose their electricity generators. Deregulation requires power generation companies to compete for customers, forcing them to operate more efficiently and cost-effectively than ever before. Coupled with the growing demand for energy created by our new information-based economy, power providers are looking for ways to do what once seemed impossible - comply with environmental regulations while operating a profitable, more productive plant.
Traditionally, reducing NOx levels in a power plant, or from any industrial boiler application, is expensive. Equipment like low NOx burners and selective catalytic reduction (SCR) can be costly, and must be installed during unit downtime or scheduled lengthy outages which can result in lost revenue for a deregulated generator. (Consider the outage time necessary for pollution abatement equipment installation and the number of plants in need of equipment under the SIP call, and watch the value of megawatt hours grow.) In some plants, even expensive new equipment may not achieve compliance, and other plants that are able to achieve standards teeter close to the edge, certain to stumble into violations during peak energy seasons like hot summers.
Power generators can reasonably expect to see NOx production fall between 15 and 35 percent with the use of optimization software products.
Reducing NOx prior to combustion by reducing the flame temperature in the boiler is often costly as well. In most cases, reduced flame temperature means sacrificing boiler efficiency. Lower flame temperatures often result in a higher loss on ignition (LOI), causing plants to generate less electricity from the same amount of fuel. According to the Electric Power Research Institute (EPRI), even expensive low NOx burners can have unacceptable increases in unburned carbon.
A Technological - and Cost Effective - Solution
A number of years ago, plant engineers and scientists began working with neural networks, a type of artificial computer intelligence based on the process of how the human brain "learns." By applying the principals of learning to the sometimes inefficient, but automated processes that produce electricity, a neural network can refine controls, improving a process until it runs at the most efficient and cost-effective rate possible for the process equipment. Neural network technology, made even stronger when coupled with other advanced computer modeling techniques, can help power plants achieve the critical balance between NOx reduction and boiler efficiency, further decreasing emission levels, and increasing the life of combustion equipment.
Optimization software tools that employ this advanced computer technology feed plant set points, biases and other operating parameters directly into a power plant's main control computer or distributed control system (DCS). In the case of NOx emissions, software tools optimize the combustion process using conventional fuel-to-air relationships, secondary air registers, and over-fire air ports to affect the fuel-to-air ratio at each burner location. With the optimizer working in this type of closed-loop fashion, it stabilizes emission levels and provides constant process tweaking to provide more consistent and lower NOx levels. Optimizers can check and adjust numerous parameters impacting NOx production every few seconds, enabling the operator to oversee and troubleshoot all other operating processes in a generation facility.
What Optimization Can Really Do - or Not Do
Power generators can reasonably expect to see NOx production fall between 15 and 35 percent with the use of optimization software products. While optimization alone may not be enough to bring some plants into compliance, these software tools provide a low-cost way to reach lower hanging fruit on the NOx tree. Optimization software, used in conjunction with other NOx reduction methods, can also create breathing room for plants anticipating the need for further pollution abatement. Through the emissions consistency gained through optimization use, plants may even see a longer life span for more expensive NOx reduction equipment.
Some optimization product manufacturers claim even higher rates of reduction - some as high as 60 percent - but power plants should take a closer look at too-good-to-be-true claims. Many available systems run "steady-state" optimization, or optimize parameters for NOx reduction only when the plant is operating at a steady rate of generation - not during a plant's naturally dynamic processes, such as startup, shutdown, load swings and dispatched operation. With steady-state systems, NOx reductions are sporadic at best, and cannot claim to be consistent, continuous, stable or effective for compliance.
Many states are again looking to fossil-fuel-fired power plants for the most dramatic NOx cuts.
Generators should also consider prior applications of a manufacturer's optimization product at similar locations. An optimization tool applied to an older fossil-fuel plant, perhaps one that has been mothballed for some time or one that has seen few modifications, may appear to reduce NOx significantly. A closer look will show the software to be part of an overall NOx reduction plan that probably includes new process equipment, other pollution abatement tools like low NOx burners, or even an entire plant equipment retrofit.
As every plant operator knows, each power plant is different. Generation in different plants is affected by a different set of variables - both internal and external. Optimization software individually tailored to meet the needs of each specific plant has a better chance of providing true NOx reduction and return on investment. Plant engineers and operators should be skeptical of optimizations systems that arrive like office software, shrink-wrapped with installation instructions, assuming each and every boiler flame operates under the same set of factors. The best optimization packages include manufacturer support for installation, service and maintenance related to generation issues and changes in the plant.
Case Study: Wisconsin Electric
In response to evolving ground-level ozone standards, Wisconsin Electric Power Company began a proactive, comprehensive strategy to prepare for new regulations and reduce NOx emissions at its Valley Generating Station in Milwaukee. Optimization software played a key role in the program.
A 300-megawatt combined heating power plant, Valley serves as a steam heat source for the city of Milwaukee in addition to producing electricity. A cycling plant, Valley is rarely baseloaded and normally on automatic dispatch. Wisconsin Electric began its efforts with low NOx burners, but, due to its dual provider role in the city, was concerned about compromising reliability when the time came to take the next step in reducing NOx.
Valley achieved a critical balance with help from optimization software which, by fine-tuning the plants automation controls, allowed Wisconsin Electric to get additional NOx reduction on Valley's number four boiler without losing efficiency.
Real-time data collected at one-second intervals and averaged over two 24-hour periods (with similar load profiles) showed NOx emissions from Valley's number four boiler decreased 15 percent by using an optimizer. Additionally, the system reduced Valley's excess oxygen by 18 percent further limiting the formation of NOx.
The best optimization packages include manufacturer support for installation, service and maintenance related to generation issues and changes in the plant.
Valley Station followed up its success with boiler four by installing optimization software on its other three boilers at the plant in 2000. Although Wisconsin is no longer part of EPA's NOx rulemaking call, Wisconsin Electric is still achieving additional NOx reductions, increasing the life of expensive pollution abatement equipment and at some point may have NOx credits to trade with others looking for environmental compliance.
Case Study: Ostrolêka Power Plant, Poland
The Ostrolêka Power Plant in Ostrolêka, Poland installed flue gas desulfurization equipment, low NOx burners and electrostatic precipitators to meet Poland's increasingly stringent air quality standards. But the owners of the Ostrolêka plant wanted to see even lower emissions, particularly of NOx. The Ostrolêka plant is located about 100 miles north of Warsaw and provides electricity for both the city of Ostrolêka and the Polish national power grid. The facility consists of three 200-megawatt units, fueled by pulverized coal.
Also using the SmartProcess Optimizer, the Ostrolêka Plant has seen consistently positive results, decreasing NOx emission levels by 15 to 25 percent while reducing LOI. The system has provided a small but measurable increase in boiler efficiency, stabilized overall boiler performance; improved response to load changes; improved steam temperature control and reduced carbon monoxide (CO) levels in flue gas.
Part of an Overall Plan
Optimization software can provide plants with a first line of NOx reductions, as well as a consistent emission rate, which increases the life span of equipment and balances the scales between plant efficiency and emission production. However, it should never be considered a complete cure for NOx reduction. As the combustion process proceeds with the software tool monitoring it, optimizers allow systems to achieve the best possible NOx levels, consistently. By placing NOx at a predictable, stable level, optimizers not only reduce NOx emissions and increase the life of all boiler process equipment, they can theoretically support a plant's ability to earn NOx trading credits.
This article originally appeared in the April 2001 issue of Environmental Protection, Vol. 12, No. 4, p. 46.
This article originally appeared in the 04/01/2001 issue of Environmental Protection.