|Kansas City Power & Light installed Siemens combustion optimization technology at its LaCygne Unit 2 Power Plant and achieved significant NOx reductions while improving the overall combustion and unit heat rate.|
By Dr. Sudha Thavamani, Siemens Energy, Instrumentation, Controls & Electrical, and Jim Stewart, LaCygne Station, Kansas City Power & Light Co.
In today’s power generation market, steam power plants are focused on identifying ways to operate more efficiently and effectively to reduce losses, maximize reliability and boost revenue. Due to continually changing demands by environmental and governmental authorities as well as consumers, plant operators and engineers are continuously seeking more effective and sophisticated technologies to support this focus and gain an edge in meeting today’s complex electricity generation market.
Intelligent combustion optimization is one method proven to accomplish plant performance improvements such as increased efficiencies, reduction in emissions, improved availability and greater fuel flexibility. Advancements in today’s optimization technologies are allowing plant operators to achieve their changing objectives while also strategically controlling plant emissions and overall operating efficiency.
Improving the Technology
Since optimization technology was introduced nearly 15 years ago, there have been misconceptions among some that combustion optimization systems are perhaps an outdated approach to increasing plant efficiency and improving plant availability. Currently, these misconceptions are being dispelled due to improved technology generating new and better results. Researchers and engineers at energy supply companies across the globe are discovering that they now have the ability to strategically assess operational data and functionality at each individual plant before the optimization process begins. This allows operators to determine the best optimization approach based on the plant’s goals.
A goal of operating at the most profitable level demands a high degree of flexibility, high efficiency, high availability and low emissions. One of the most cost-effective and common ways to improve boiler efficiency is to apply primary measures based on optimized combustion adjustments. However, this optimization approach is usually restricted by the power plant operator’s limited knowledge of actual combustion conditions. These uncertainties about the actual combustion process lead to situations where operators keep most boiler settings constant, although considerable variations occur with respect to fuel properties, fuel flow rate imbalances, load range or air flow disturbances.
In general, insufficient monitoring and control means that the operation of the boiler is based on the use of certain combinations of global or indirect variables, derived either from the recommendations of the boiler manufacturer or from the accumulated experience of the operators of the specific plant. Combustion optimizer technology was designed to give operators a better understanding of the plant’s data and how to prioritize it to increase overall efficiency, but it’s the proper application of that data that makes or breaks the optimization process. Today, more sophisticated technologies exist that allow software providers to work with operators at the onset of the problem, looking at data differently to produce better, faster and smarter result to combustion variables.
Variables in Application
Though initially well received, early optimizers began exhibiting systematic issues and operators began turning them off. Today, the importance of these tools is being realized and efforts are underway to improve the older technology and provide engineers and plant operators a reason to trust the systems once more to improve heat rate, lower emissions and improve availability. Newer optimizer systems feature more advanced controls, control more data and are more adaptable to changing conditions. Simply stated, newer upgrades feature a carefully constructed “recipe” emphasizing only the most relevant data and minimizing the rest.
Modern combustion optimization technologies provide closed-loop optimization of fuel and air mixing by manipulating fuel and air levels to balance combustion in the furnace. They are software-based solutions that require no more than minor modifications to mechanical equipment and are relatively straightforward when it comes to operations and maintenance training. Each system is comprised of measurement systems and combustion optimization controls. Some companies, such as Siemens Energy, use modules for laser-based measuring technology, distribution calculation based on computer-aided tomography (CAT) procedure and combustion optimization controls based on both mathematical modeling and neural networks.
But the tools are only part of the trade. The key factor in promoting cleaner, more efficient power plant operations while still leaving space for more flexible operation is to truly take into account the appropriate recipe of data and control to properly determine the best optimization approach.
Major strides have been made in this realm. Offering qualitative and quantitative audits of system controls at the onset of the problem was discovered to be the best way to ensure plants are producing the correct data to determine the best application for the use of technology. For instance: if a plant operator is aware of a very specific operational problem – such as slagging, derates or heat rate gain – and has data that can pinpoint it, he or she can rely on classic optimization controls to alleviate the issue.
However, immediate availability of such precise data is rarely the case. When little can be directly measured on a plant operation, more advanced solutions are needed to help solve what is a far more complicated problem. In this situation, laser-based or even hybrid measurement systems in unison with the appropriate “closed-loop” combustion optimization solutions will produce the maximum achievable benefits.
Applications in Action
In recent years, a growing number of modernized combustion optimization success stories have been surfacing around the world, delivering renewed and warranted faith in the technology. From the U.S. to Germany and to China, plant operators are realizing the benefits of using intelligent combustion optimization technology to fix specific problems or even to ensure their plant is running more cleanly and efficiently.
Application 1: Lowering Coal-Fired NOx with Robus Hybrid Optimization
In 2011, operators at Kansas City Power & Light’s (KCP&L) LaCygne Unit 2 Power Plant were looking for a way to reduce stack emissions of NOx. This 30-plus-year-old 720-MW balanced draft, Carolina-type, B&W wall-fired unit has seven MPS-89 pulverizers, each having eight coal outlets that make up an individual row of burners. The burners are second-generation, dual-register, low-NOx burners. There are seven rows of burners, each with a compartmental wind box that has two controllable dampers. The unit did not have an overfire air (OFA) system.
LaCygne operators turned to Siemens to address their concerns. In this instance, the optimization process used a hybrid structure of closed loop controls with neural network optimization technology, integrated with a laser-base combustion measurement system.
The integration process began first with the installation of the laser measurement grids inside the combustion chamber. Next, parametric testing was done on the conditions of the boiler – followed by deduction and analysis of spatial distributions – to determine the required controlled variables for the closed loop controls engineering. Finally, the Siemens combustion optimization control system was integrated into LaCygne’s existing plant DCS System.
Model Based Controls Target Emission Reductions under Strict Operational Restraints
The laser-based measurement system maps the concentration of in-furnace CO, O2, H2O and temperature simultaneously in real time and directly in the furnace. Laser transmitters and receivers are arranged outside the boiler resulting in a grid of laser beams crisscrossing the furnace. Each path measures an average value for temperature, O2, H2O and CO simultaneously, and together the paths are used to create a tomographic image of this plane in the boiler that is also displayed to the operators directly in the control room.
Temperature and concentration distributions are calculated from the measured path averages with the aid of the CAT procedure. CAT algorithm was used for calculating certain characteristic distribution data, e.g. values at different grid points, averages, minimum and maximum values for different paths in the distribution, skewness, etc.
The laser measurement system is integrated into the Siemens Optimization System Server, which is in turn connected to the ABB Infi90 Plant DCS. These connections permit access to measured data from the existing DCS, such as coal and airflows or CO and NOx in the flue gas, by the combustion optimization process and also facilitated the optimization of process signals back to the DCS.
A watchdog signal allows communication verification between the optimizer and the DCS. The process measurements from the DCS along with the laser measurements data from the boiler are used to calculate the optimized values within the Siemens optimizer. These values are integrated into the plant DCS as biases to the existing setpoints in the DCS, which are transferred to the basic underlying controls and to the field.
Safety measures are followed in sequence for incoming setpoint biases before they are used in control and boiler processes. The operator can smoothly switch the optimizer controls with the ON/OFF switch. If the optimizer loses connection to the DCS, the control interface design automatically transfers control back to the plant DCS and control room operators without introducing a transient condition.
Distributing air properly is a critical step in combustion optimization. The objective behind CO balancing is to uniformly spread the CO distribution across the boiler. In wall-fired units, the countermeasures involved in lowering NOx include air and fuel staging. This methodology reduces stoichiometry during combustion and minimizes the formation of both unburned fuel and thermal NOx, improving the completion of the combustion process in the boiler. Hence the staged combustion helps in better completion of the combustion process in the boiler. At LaCygne, the staged combustion technique for NOx control involved the re-distribution of secondary air (staged air) from the main combustion zone to the top levels while reducing air in the bottom levels of the furnace. This moves the compartment air to the “air-starved” regions in the furnace, and improves the air/fuel mix within all regions.
Neural Networks Control Multiple Variables
Neural network technology can be viewed as a multivariate nonlinear non-parametric estimation tool. It shares a descriptive term from biology in that they are represented as networks of simple neuron-like processors. This highly adaptive technology uses a unique combination of neural network and complex systems algorithms to learn the complex interactions of process variables from historical data.
The manipulated variables for the neural model used at LaCygne Unit 2 were secondary air north/south damper positions, compartment airflow setpoint and feeder speeds. Some of the input variables comprise of laser measurement readings of temperature, O2 and CO profiles, and unspecified basic conditions such as load. The output variables include the stack CO measurement. The primary target value for LaCygne was the Stack NOx.
With the help of the neural network, the optimum manipulated variables for the specified target of NOx reduction are determined with due consideration of constraints. The constraints were related to maximum allowable stack CO and opacity. The other modeling constraints were to keep the total air and the sum of feeder speeds constant, at a given load condition.
With the use of the SPPA-P3000 Combustion Optimization technology, LaCygne saw significant NOx reductions while improving the overall combustion and unit heat rate.
Additional benefits of using Combustion Optimizer at LaCygne Unit 2 Power Plant included:
- Enhanced and balanced combustion
- Significant improvements in boiler slagging
- Better CO distribution in boiler
- Supplementary O2 reduction based on the balanced combustion
- Improved unit heat rate
- Better centralization of the fireball
- Staged combustion achieved emission reductions
These improvements were attributed largely to the combustion balancing controls, fuel/air staging controls and O2 reduction controls. Coal and airflow staging control solutions for controlling the distribution of air and pulverized coal flow to individual boiler level resulted in lower emissions. O2 setpoint reductions proved to boost the benefits of staged combustion producing even lower NOx emissions.
Application 2: Enhanced Plant Performance with Combustion Optimization
A SPPA-P3000 Combustion Optimization System was also installed almost identically in 2011 at Unit 4 of Huaneng Rizhao Power Plant, P. R. China and has already boasted significant performance improvements. This unit is a 660-MW tangentially-fired, supercritical boiler using soft-coal fuel. To balance the combustion, a closed-loop optimization is delivered by manipulating fuel and air levels with the help of laser measurements, which provide the key combustion components O2, CO, H2O and temperature simultaneously, directly in the furnace.
Parametric Testing Leads to Improvements in Boiler Behavior
During the parametric testing phase, the behavior of the Rizhao boiler was observed by changing and adjusting individual secondary air, secondary auxiliary air, and SOFA dampers. The testing and combustion analysis showed that the unit’s fireball was often not in the center of the boiler where it belonged, creating an unbalanced heat transfer to water wall and heaters. This also increased the production of slagging for parts of the water wall. Testing also confirmed that the non-homogeneous distribution of excess air in the boiler led to efficiency loss in regions with too much air and bad combustion in regions of starved of air. The combustion optimization controls at Rizhao allowed for a number of situational improvements in the boiler to improve efficiency. Secondary auxiliary air dampers helped to center the boiler’s fireball for improved heat transfer and reduced slagging. Improved control of SOFA dampers facilitated a proper O2 balance for more uniformed distribution, reduced O2 and lower emissions. Secondary boundary air dampers created an improved air-to-fuel ratio for more uniform combustion.
The fireball centering, O2 distribution and combustion balancing controls each helped to allow the reduction of the excess O2 for increased boiler efficiency overall. This additional logic using CO concentration was implemented to determine the O2 setpoint correction dependant on the actual combustion situation in the boiler. The CO values in the furnace measured by the laser measurements, and the CO after boiler from the DCS, were used to determine the rated CO concentration. The CO setpoint to the integral control was determined as a function of the unit load. The optimization controller was restricted by the lower limit for O2 reduction, which was based on the function of unit load and had an absolute lower limit from the O2 setpoint characteristics from the DCS. A reduction in O2 was achieved when the optimizer was switched ON.
An Improvement in Overall Efficiency
The combustion optimizer was tested for different plant conditions like that of different unit loads, different coal etc.
Additional results of using this optimizer were clearly discovered:
- Better centralization of the fireball
- Better O2 distribution in boiler
- More O2 reduction based on the balanced combustion, normally the O2 reduction is 0.7~1.1
- Reduced auxiliary power, reduced coal consumption and increased boiler efficiency
At the request of the management at the Rizhao Power Plant, a third-party evaluation of boiler efficiency improvement achieved by Combustion Optimizer was administered by Xi’an Thermal Power Research Institute Co, Ltd. (TPRI).
TPRI conducted a boiler efficiency test using GB 10184-88 standard methodology at full load and at partial load to compare the boiler efficiency with combustion optimizer ‘OFF’ and ‘ON’. Performance improvements attributed to combustion balancing but not included in the TPRI boiler efficiency calculation such as lower auxiliary power requirements (from induced, forced and primary air fan loads) and increased steam enthalpy were also calculated using TPRI test data and ASME 4.1 standards.
The combustion optimizer improved the performance by a total of 0.57 percent at full load. A more balanced combustion was achieved permitting the excess O2 to be reduced by 0.93 percent based on the DCS setpoints. This resulted in the decrease of fan loads (induced/forced draft and primary air), which further diminished the auxiliary power requirements by 293 KW/hr, or 0.05 percent of the load. This reduction in auxiliary power led to the increase in the amount of electricity available for sale by the plant.
Also, at full load the steam enthalpy increased as superheater and reheater steam temperatures were enhanced as a result of better combustion which increased the steam turbine output by 0.07 percent overall. In addition, NOx levels were reduced by 14.4 percent as a result of lower excess O2 while the CO and LOI did not materially change.
Revalidating the Need for Optimization
The experiences at Unit 2 in LaCygne and Unit 4 in Rizhao have shown that the appropriate use of real-time combustion optimization technology in a coal-fired plant can definitively improve combustion and heat rate while reducing emissions and the potential for lost generation. Both plants experienced documented, improved efficiency through optimizing their combustion. If the technology is properly designed and applied, plant operators and engineers can make appropriate adjustments to improve unit performance with minimal training and eliminating the need for costly mechanical changes. Improvements to combustion optimization through the years have validated a second look at the software-based tools as a means of improving plant efficiency and lowering emissions.
1. Boiler Efficiency Calculation via ASME PTC 4.1 Method
2. GB 10184-88 Performance Test Code for Utility Boiler
Power Engineerng Issue Archives
View Power Generation Articles on PennEnergy.com