Power Plant Restart

Power Plant Modeling: Restart with Confidence and Lower Risk During Emissions Control Revamp
An Industry White Paper

By Gary Hida, Applied Control Engineering, and Mark Gallant, Aspen Technology, Inc.
2011 Aspen Technology, Inc. AspenTech, aspenONE, the Aspen leaf logo, OPTIMIZE, and the 7 Best Practices of Engineering Excellence are trademarks of Aspen Technology, Inc. All rights reserved. All other trademarks are property of their respective owners. 11-1055-1111
About AspenTech
AspenTech is a leading supplier of software that optimizes process manufacturingincluding oil and gas, petroleum, chemicals, pharmaceuticals and other industries that manufacture and produce products from a chemical process. With integrated aspenONE solutions, process manufacturers can implement best practices for optimizing their engineering, manufacturing and supply chain operations. As a result, AspenTech customers are better able to increase capacity, improve margins, reduce costs and become more energy efficient. To see how the worlds leading process manufacturers rely on AspenTech to achieve their operational excellence goals, visit www.aspentech.com.
Introduction
Today's new emissions control systems are very complex, challenging E&C firms with their design and construction, as well as owner/operators with their control, operation, and maintenance. The following document discusses a 50-year old coal fired power plant that was mandated to reduce emissions or lose their operating permit. They needed to reduce the mercury emissions and sulfur dioxide emissions by over 95%. To do this they contracted a supplier to install a new wet Flue Gas Desulphurization (FGD) unit. These new environmental units are increasingly complex and operate quite differently than other power plant equipment. The plant operations and engineering teams wanted to better understand the operation of this new equipment and how it would work within their existing infrastructure. A single day of lost production would cost $200 to $400 thousand dollars, so avoiding any unplanned outages was a critical concern. In order to reduce the risk of failure and increase safety, they evaluated different scenarios for power plant restart using dynamic simulation. A team was assembled with members from the plant, the E&C firm and a US-based process control engineering firm. They worked to create a model-based dynamic simulation of the plant with the new equipment and ran a number of simulation cases. They tested various startup and shutdown sequences, plant cycling, as well as a number of disturbances and trip conditions. They then used this data to create the control logic and operating procedures long before the equipment was even installed. We will discuss the model setup, execution, data output analysis and best practices developed.
Power Industry Dynamics

The power industry is facing many dynamic pressures today. Some markets are being deregulated, organizations are diversifying their fleet of assets and there is the continued challenge of retaining and developing the talent pool. As the quote states below, change is now the norm for the power industry. One of the intriguing challenges today is the unprecedented regulatory pressures around environmental controls and constraints.
More change will happen in our industry in the next 10 years than has happened in the last 100.
San Diego Gas & Electric
Companies from around the world are spending billions of dollars on emissions control systems. Southern Company has published that in the last three years alone they have spent $3.1B on emissions systems to improve their fleet. This is not a unique situation, as attested to by the annual reports of many companies in the Power industry. Many of these companies are allocating a large portion of their capital budgets toward environmental controls. These new environmental systems are creating additional complexity into the power industry infrastructure, challenging the operational norms and changing industry best practices and organizational focus. If you look at these systems, they look more like a mini chemical plant than a power plant. Chemical processes are controlling those units, posing a challenge to power operators.
New Emissions Control System

In this case the plant was about to receive a major new emissions control system. This was an old plant with very old boilers, and the owner required that the FGD unit operate smoothly within the existing infrastructure and meet their mandated safety and reliability operating standards. As this was a completely new system, determining how it is going to work and how they were going to control it was critical. They were, after all, dropping in an advanced system into an old facility that has been operating the same way for many years. Adding more systems, with good-sized booster fans and more control loops, created complexity. The plant owners objectives were to understand these key issues: 1. Sizing - Validate that the equipment specified by the supplier at a cost of 10s of millions of dollars was in fact sized appropriately. 2. Control How will the unit work and the proper control strategy. This supplier had never installed the proposed FGD unit on a plant with this unique boiler and control system configuration. 3. Start up How to start the unit for the first time, impact of new or seasoned bags in the bags house on future startups, and best sequencing given different operating modes. 4. Operation - How to start, operate and shut down the unit. How will the unit respond when cycling up or down to match grid signals? 5. Risks How stable will the plant be? They needed to understand, identify and reduce the operational risks. What happens when certain components trip or fail? 6. Training Can we train our operators on the unit before it is installed to document best practices and update all our procedures? Finally, how was the installation going to impact the other planned projects during the shutdown? The plant was going through a major revamp of the turbine and supporting systems at the same time. Could modeling help them answer all these questions and give them critical path time back during the shutdown?
Approach
Typically, the approach to the concerns faced by this power company would be to design a control system after the process has already been fixed and the systems are in place. The next step would be to train operators using standard operating procedures, discovering operability and control issues during the start up and commissioning phase. This requires a longer or extended start up phase so that they can really tune the system. The challenge with this approach is that it can result in inherent unit operability and control problems. This in turn can delay the start up of a plant and leave operations personnel blind to underlying safety issues. The possible cost of failure in this typical approach led the power plant to undertake the dynamic simulation approach. Dynamic modeling and simulation can drive many benefits as opposed to the traditional approach. For one, integrating the design and control system early in the design phase helps pre-tune the control loops and provide a better understanding of its stability, enabling the operator to easily make changes before the systems are live. The operability / engineering studies help you understand the dynamic behavior and response of the new equipment. For example, instead of just understanding how the system runs in steady state for a given load, you want to understand how it is going to operate when you are cycling the plant up or down. You want to investigate and test all of the transitions to
see how the unit will operate. This is especially critical as they were all concerned with disturbance handling and various trip events. Safety is of top concern, so looking at the plant systems through dynamic simulation was key to make sure everything operates correctly at startup, and to see how the plant is controlled and how the plant can be shut down safely and reliably without any hazards. Static models just do not allow you to understand all these dynamics together. Finally, dynamic simulation allowed the plant to train the operators. These new systems are complex, potentially operate differently and create new challenges. Having the operators react to the challenges in dynamic simulation environments allows for best practice approaches to be uncovered, developed and formalized during all phases of the plant operations.
The Process
Develop the Static Model
The plant owners chose Applied Control Engineering (ACE) as their modeling vendor. ACE collaborated extensively with the engineering firm and the plant engineers to collect the data to model the complete unit and all the various gas flow paths. The engineering firm had already started a static model and assembled much of the data. Using a commercially available modeling tool, Aspen HYSYS Dynamics from AspenTech with its extensive logic library, ACE took the EPCs process flow diagrams and built the static model, then demonstrated the model at various plant loads (See Figure 1).
Figure 1: Develop the static model.
The display in Figure 1 shows that while building and running the model, it can be customized to show the exact level of detail required for the user. In this case, the material balance spreadsheet is displayed along with the process and control model. This can make it very simple to verify various operating points.
Create the Dynamic Model

Once all the information is entered and the static model is verified, ACE converted the static model to a dynamic model. As previously noted, a dynamic model is of much more use to a controls engineer for examining possible control strategies and for operator training. The selected modeling tool made the conversion from static to dynamic model simple and completely seamless. Additional control elements and strategies were then added to enable the various simulation modes (See Figure 2). This included binary switches that allowed disturbances and other factors to be turned on and off while the model was running. For example, this model had switches which allowed demonstrating the loss of the fluidized beds, starting and stopping of the 3,600hp booster fan, as well as the two most important transient conditions that the plant was interested in examining initial startup and unit cycling.
Figure 2: Control elements and strategies added.
The basic process flow diagram (PFD) display can also display real-time process conditions as shown in Figure 2. So in real-time or model-time, the power plant engineers and operators could see how various process conditions were affected by the simulated scenarios and various disturbance modes.
Model Scope, Features, Assumptions and Limitations

To meet the stated goals of this project, the requirement was to model all gas flows, starting upon boiler exit through the new booster fan, the FGD unit, bag house all the way through the stack, including all existing fans and structures. Specifically included in this model were all the controllers, variable dampers and 5 fans controlling gas flow, the 2 existing FD fans, the 2 existing ID fans, and the new 3,600hp booster fan. The majority of the equipment modeling and control strategy development was around this equipment. As we know, any model is only as good as the information that is used to build it. The selected modeling tool provided an extensive physical property database which was used to validate and insure that the model was accurate. Given the scope of this project, the modeling tools chemical reaction modeling features were not required and were not used. The dynamic simulation model created enabled the engineers and plant personnel to understand how the equipment would react once installed. As noted, the start up, shut down and trip dynamics were the key points of concern as the booster fan would be started with the plants in full operation. The simulation covered nearly all operating modes from minimum to maximum load and focused on the transient conditions (See Figure 3).
Figure 3: Operational modeling in Aspen HYSYS.
Simulation Scenarios
A broad series of simulations and scenarios were run. The key points that were examined during each were the gas flow, the pressure, the fan horsepower, and the operating points. The objectives of the simulation scenarios was to see if recycle of the flue gas was going to be needed and what the effect would be of clean bag or seasoned bags in the bag house. The first scenario was steady state with the FGD unit offline. Next, they evaluated how everything would respond and operate during the very first cycle. Finally, they studied the start up of the booster fan during a variety of plant loads. This work provided valuable information to the engineers, confirming the static model and giving very important dynamic information about these transitions (See Figures 4 and 5).
Figure 4: Flows during booster fan startup, without recycle.
Figure 5: Pressures during booster fan startup, without recycle.
Other transient conditions, including cycling plant loads, were examined on the overall stability of the unit and its impact of the selected control strategy (See Figures 6 and 7). At the bottom of the trend you will see a multi-colored bar; this is a time display showing how far along in the simulation the data is being displayed.
Figure 6: The response of fan horsepower.
Figure 7: Flows during rate changes 50% to 75% to full power.
Finally, they examined what happens when the new booster fan trips with the plant in full operation. The model suggested that the furnace pressure would initially go positive, but that if the bypass is opened then the plant could operate safely with only a slight de-rate output.
Conclusion
The project was a great success. The unit started smoothly the first time because of the multiple scenarios tested using a dynamic simulation approach. The smooth start was even more remarkable because a two-week delay in the turbine rebuild resulted in only one hour of booster fan test time prior to start up. Using the dynamic simulation modeling tool, Aspen HYSYS Dynamics, along with input from all the members of the team, they were able to answer each business challenge. The FGD unit and booster fan was the right size, they understood how it would work, how they will control it, what issues might be seen during start up, how to operate the new asset, and they validated that the plant would in fact be stable while fully meeting all operating, safety and reliability standards. A detailed technical report was produced which described all of the cases and scenarios that were run, as well as documented the issues that were expected to be seen in the control strategy. Data collected post commissioning was compared to the modeling data and found to be extremely close, further validating the results and the process selected. As a direct result of the modeling, several severe control strategy problems were identified and critical refinements were developed and further tested before the unit was installed. The plant has not experienced any issues in operations due to control and operation of the FGD unit. ACE and the power plant team were able to complete the modeling project under budget and weeks ahead of schedule, and most importantly, significantly lower the startup risk for the power plants owner. With the new emissions control system in place, the coal power plant has gone from one of the states largest polluters to one of the countries cleanest!
References and Articles

Evely, Kosowski & Kosak (2008), Improving Control System Design with Dynamic Simulation, ISA EXPO 2008. Gas Compressor Performance Website (2011), www.mycalculations.com/chem/jmm/chem16. Authors referenced on website: Coker, AK; Fortran Programs for Chemical Process Design, Analysis & Simulation; Gulf Publishing Company, Houston, 1995. Chopey, NP; Handbook of Chemical Engineering Calculations; McGraw Hill, New York, 2003. Siemens Energy & Automation, Inc (January 2005), Combustion Management Solutions Furnace (Draft) Pressure Control, http:/ /cache.automation.siemens.com/dnl/zQ/zQzNjI5NjcA_51436060_HB/AD353-106r2.pdf, AD353-106. Marlin, T., (2000). Process Control: Design Processes and Control Systems for Dynamic Performance, 334-337,426-427. McGraw Hill, New York. Morrison, T., (2006). Boiler Case Study: Minimizing NOx without Sacrificing Efficiency. Available online: http:/ /texasiof.ces.utexas.edu/PDF/Presentations/Apr3_08/SterlingChemicalsBoilerCaseStudy. June 2010.
Worldwide Headquarters
Aspen Technology, Inc. 200 Wheeler Road Burlington, MA 01803 United States phone: +17812216400 fax: +17812216410 info@aspentech.com
Regional Headquarters
Houston, TX | USA phone: +12815841000 So Paulo | Brazil phone: +551134436261 Reading | United Kingdom phone: +44(0)1189226400 Singapore | Republic of Singapore phone: +6563953900 Manama | Bahrain phone: +97317503000
For a complete list of offices, please visit www.aspentech.com/locations

Power Plant Restart

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Power Plant Restart

Uploaded by

Copyright:

Available Formats

Power Plant Modeling: Restart with Confidence and Lower Risk During Emissions Control Revamp

An Industry White Paper

Power Industry Dynamics

New Emissions Control System

Figure 1: Develop the static model.

Create the Dynamic Model

Figure 2: Control elements and strategies added.

Model Scope, Features, Assumptions and Limitations

Figure 3: Operational modeling in Aspen HYSYS.

Figure 4: Flows during booster fan startup, without recycle.

Figure 5: Pressures during booster fan startup, without recycle.

Figure 6: The response of fan horsepower.

Figure 7: Flows during rate changes 50% to 75% to full power.

References and Articles

For a complete list of offices, please visit www.aspentech.com/locations

You might also like