Improving Energy Efficiency & Reducing Cost in Utility Systems
/With the processing power of computers following Moore’s Law and utility plants becoming more and more complex (steam at several pressure levels, varying fuel diet, cogeneration to supplement electric power import or to sell electricity, turbine motor swaps and emission limits, etc), it is inevitable that computer based programs are now used to simulate and optimize plants to improve energy efficiency and reduce energy cost. One such computer based software program is Visual MESA which can be built to precisely simulate and optimize plant utility systems with exceptional levels of thermodynamic accuracy. The software uses mixed-integer sequential quadratic programming (SQP) optimizer as it is well suited for solving problems with significant non-linearity and discrete decisions (i.e., start-stop of certain pieces of equipment) which is the characteristic feature of utility systems. This article shines light on how the software can help the plant operators to improve energy efficiency and reduce energy cost of utility system operations.
Visual MESA Capabilities
The Visual MESA software uses the following five capabilities to improve energy efficiency and reduce energy cost in utility systems:
- Modeling
- Optimization
- Monitoring
- What-If Case Study
- Results Historisation
1. Modeling
fig 1: steam imbalance of a high pressure header in a typical plant
The exact condition of the plant can be modeled with the details provided and can be configured in the software. The typical details required would be equipment specifications, demand and supply of the plant and costing of over the fence purchased/sold utilities. The data is made live by connecting the model to a real time database. This helps the user understand the energy and material balances of the plant and the price of the utilities in an explicit way and also some of the limitations which are sometimes hidden. The software provides a balloon as a representation for imbalances for each steam, fuel and electrical header and the user can find out any imbalance based on the size of the balloon and take corrective actions for faulty instruments, inaccurate design specifications and missing supply or demands. Also, the split up of cost for each utility is provided for the items to be targeted for maximum or minimum use. For example, figure 1 shows the steam imbalance of a high pressure header in a typical plant.
2. Optimization
The software is a real time optimization application using a nonlinear mixer-integer SQP (Successive Quadratic Programming) optimizer and the optimization objective function of the program is to minimize,
Total Energy Costs = Fuels + Electricity + Other Costs
The objective function is achieved by using the optimization actionable variables (type of fuel to the boilers, pump swaps, gas turbine loads, electricity import or export, multiple extraction/condensing turbines flows, steam let down and vents) subject to constraints (burner capacity, operational, contractual and emission limits).
An example of the optimization of the minimization of fuel addition to one of the boilers is shown below.
The optimization can be open loop or closed loop depends on the preference of the customer. The difference is that in closed loop the optimizer’s recommendations are automatically implemented on continuous but not discrete variables whereas in open loop it serves as an advisory role to operators to implement them. The difference between the simulation and optimization is the savings. The figure below shows the representation of savings in Visual MESA.
fig 2: savings per hour after implementation
3. Monitoring
The Visual MESA system is presented on a web browser environment, providing a variety of monitoring features for the entire fuel, steam, boiler feed water and condensate system and highlights problems. These includes trending data, alerting of big changes, validation of sensors, motor and turbine status for swappable pumps, etc.
For example, trending a steam flow sensor of a boiler can help the operator visualize the behavior of flow from a boiler over a period of time and can figure out abnormalities in the trend, if any. The monitoring feature can also be used to schedule maintenance of an equipment based on its performance, say for a boiler when its efficiency goes down it can be due to variety of reasons affecting economy of the plant. The given figure below is an example of trending of a sensor in Visual MESA,
fig 3: sensor trending in visual mesa
Since all plant users do not need access to the Visual MESA software model as a whole (i.e. have it installed in their system) an Excel based custom report is added-in with the software to provide a familiar environment for users to view information about the complete energy systems of the plant. This serves as another monitoring feature to share data especially for people at top management levels who are interested in just a few plant key performance indicators. Report can be published as a web page and saved as a pdf file for easier distribution and access.
The figure below shows an example of a report generated with Visual MESA. In the worksheet shown, steam generation and fuel consumption are reported for the current values and the optimized values, with delta being the difference between them, can be increased, decreased or left alone depending on the existing scenario.
figure 4: visualmesa report
4. What-If Case Study
A what-if case study can be used in Visual MESA to help the user evaluate how the steam system will respond to a small or major changes. The change can be in fuel pricing, additions or removal of an equipment, performance enhancement of equipment, emission limits, etc - using current, historical or user defined data. It can also be used to plan or train for upcoming events such as maintenance or repairs of equipment. Economic savings are found by comparing a given base case to the optimized case that includes the proposed change and it is important that everything in the two data sets are same except for the change that is being studied. This can result in the plant operators being able to plan optimum plant operation, thereby minimizing cost and maximizing savings.
The base for the what-if case study can be either the current system operation or it can be gathered from historical data, which are easily accessible using the standard software tools, which allows retrieving current or historical data from the real time database.
5. Results Historization
Visual MESA can historize the selected results of the real-time database and retrieve whenever it is needed for analysis and to identify areas of improvement, and to be presented in an associated dashboard. The parameters most commonly historized are efficiencies of boilers, furnaces and turbines, gas turbines heat rate, header imbalances (steam, fuel, electrical), pump and motor availability statuses, process plants total and specific energy consumption, and whatever else the user feels is important. The historical data can even be extracted from Visual MESA using the excel add-in function to an excel sheet for current value and historical value and can be saved as values only to be communicated to other interested parties.
Conclusion
Intelligent systems are here to stay. Visual MESA is one such system that is robust with a mature and industry proven solver that can converge within a reasonable time and help plant personnel to handle the plant in a more efficient, economically and profitable way on a real-time, routine basis.
Click here for more articles on Visual MESA.
