Ken Parker, Dr Norman Plaks & Dr Anupam Sanyal, International Environmental & Energy Consultants Inc, UK & USA
For particulate pollution control, most coal fired generators are fitted with electrostatic precipitators (ESP), the operational performance of which can be compromised by the firing of low sulphur, high ash coals of countries like India and China or high moisture, low sulphur from the Western USA, resulting in emissions well in excess of the host countries’ legislation.
Although the performance of existing ESP can be increased using traditional methods, such as increasing the collector plate area or by flue gas conditioning to overcome the reverse ionization problems, these approaches are expensive. To mitigate some of this expense two novel computational methodologies have been developed.
The PALCPE (Proactive Approach to Low-Cost Precipitator Enhancements) methodology incorporates well recognized mathematical physics that describe the operation of the ESP, and which are duplicated in the algorithms of the powerful ESPVI 4.0W model. This model, originally developed by the USA Environmental Protection Agency, which when synchronized with the full-scale ESP accurately duplicates its performance; for all-intents-and-purposes, the model is the precipitator and the precipitator is the model.
Particle collection in a precipitator is best described by the Deutsch Equation (DE), which in its scientifically rigorous form is applied to the model. The model’s ability to duplicate space charge effects and the use of DE are the keys to combining ESP experience and knowledge with expert modelling in order to perform any enhancement analysis.
All essential precipitation processes are duplicated within the PALCPE model. The diffusion and field charging continuum is derived by dividing the size distribution into 27 individual increments, for each of which the space charge contribution is computed. The continuing decrease in particle concentration, from inlet to outlet, is achieved by dividing the precipitator into individual length increments, upstream and downstream of each electrode element, where the incremental space charge is computed.
After the precipitator’s physical size and gas flow properties are entered into the model, the electrical characteristics of the discharge electrodes are determined by matching actual ESP generated air load VI curves to model generated ones. Air load data are primarily used to isolate the electrode’s electrical characteristics from the particle space charge effects in order to set the model electrode characteristics. It is this matching of the site clean air curves, which enables, whatever actual format is used, to be represented by a simple round wire or array by the model. In the case of precipitators from the Newly Independent States (NIS), the Russian electrode format takes the form of a strip having protrusions or needles punched out, such that the needles are at right angles to the strip. Figure 1 indicates how the characteristics of this electrode can be represented in the model, either as a single diameter electrode, or an array of elements.
The next operation is to determine the ash particle size and concentrations by carefully comparing the ESP’s individual VI curves under actual operating conditions to ones generated by the model, using the now fixed DE data file. The identifying key is the change in space charge corona suppression, for the full-scale to model generated curves for differing particle property settings. When the curves show reasonable agreement, the model’s particle settings represent those entering the precipitator.
All possible ESP performance changes can ultimately be reduced to three variables – the collection area, the gas flow rate, and the electrical migration velocity, which is the terminal velocity of a charged particle in an electric field resisting the viscosity of the gas in which it is suspended. Expert knowledge and experience can then be applied to explore the effects of changing the three variables using proven performance enhancements to both the ESP and combustion processes. Low cost performance enhancements could include changes to the energization system, substituting discharge electrodes that improve electrical operation and tuning the combustion process to decrease excess air and/or provide more favourable temperatures to decrease ash resistivity.
The development of PALCPE came about from the pooling of experience and knowledge gained from working on projects to enhance the performance of older and smaller precipitators in the NIS.
However, PALCPE requires the precipitator to be in a satisfactory operating condition to gather good quality and accurate electrical operating data. All too frequently the data are sparse and incomplete, which prevents PALCPE direct application.
A paper, presented at the IX International ESP Conference in May 2004 – “Practical Experience and Results of ESP Operation Collecting High Resistivity Fly Ash with High Dust Load” by A. Rustambayav, A. Koptov, K. Porle), however, provided an opportunity for validating a method that would eventually become VESPR (Virtual ESP Reconstruct).
VESPR calls upon experience and knowledge to fill in the missing input data, so that it seamlessly matches as much as possible the actual operation of the full-scale ESP. This means that instead of using data from a poorly operating precipitator, all work is done with a virtual ESP. Once the VESPR work is satisfactorily completed, PALCPE can then follow to investigate all potential enhancement scenarios.
The paper described the rebuilding of a 1960 design East German LUK ESP at the Aksu power station in Kazakhstan by Alstom, which unlike many precipitators, used different field lengths. The only original portion that was retained was the foundations, which gave the rebuilt unit the same footprint and field lengths. The collector plate height was increased from 12 m to 15 m and the collector spacing was changed from 300 mm to 400 mm; while the original electrodes were replaced with spiral electrodes.
The VESPR rebuilt precipitator made use of physical data from a LUK design precipitator at Reftenskaya, VI data were available from Troitskaya, which fired the same high ash Ekibastuz coal, while Spanish spiral electrode air load data was used to set up the DE data file. The virtually constructed rebuilt ESP indicated an operating efficiency of 99.64 per cent. This corresponded well with the reported measured efficiencies of 99.6-99.7 per cent. Once the virtual precipitator was operating, PALCPE was able to demonstrate how the performance could be further enhanced.
PALCPE has been successfully used to upgrade the ESP performance at three Russian thermal power plants, which fire high ash, low sulphur coals producing a high resistivity ash, and hence high particulate emissions. In addition to improving the precipitator’s mechanical condition, further enhancement has been concerned with increasing the migration velocity using improved electrics to keep operation out of the back corona regime, which has resulted in a total emission reduction of close to 110 000 tonnes per annum.
VESPR has been used to analyze an Indian precipitator, which also fires a high ash, low sulphur coal producing a high resistivity ash. This 100 MW seven-section precipitator has a design specific collection area of 160 m2/m3/s and a design efficiency of 99.37 per cent. Figure 2 shows the actual site VI data that were available. Using VESPR principles, the precipitator was reconstructed within the computer and optimized electrical operating conditions were developed for each of the seven sections as indicated by the points labelled Modelled 1A to 7A.
Figure 2. Application of VESPR electrical operation to the Indian precipitator
The virtually reconstructed Indian precipitator was determined to be operating at the design efficiency. It should be noted that the VESPR derived electrical operating points (Figure 2) are all located in the regime below, which the original VI curves indicated back corona. Precipitators collecting high resistivity ash should, for best performance, operate below the onset of back corona. Once VESPR was completed, PALCPE was able to demonstrate further performance improvements that could be obtained to meet emerging emission requirements by the application of relatively low to modest cost proven scenarios.
Correct usage of these approaches brings any ESP up to its optimum capability within the computer, virtually and very inexpensively. This contrasts with conventional upgrade approaches where the precipitator is first brought up to its optimum performance, which may then require some of the newly installed replacement parts be replaced in favour of more advanced components to meet the performance enhancement required.
The ESP is an expensive unit of equipment for which the choice of the optimal enhancement option should not and cannot be taken lightly. A poor decision can be excessively expensive. A plate area extension on a 500 MW unit can cost several million dollars, as against much less investment for proven techniques, such as improving controllers, optimizing discharge electrode characteristics and reducing gas volumes.
PALCPE augmented by VESPR identifies these and other upgrade options that are available quickly and economically. The cost for applying computational analysis is minute in comparison to the high cost for potentially less-than-optimal, conventional upgrade enhancement approaches.
PALCPE successfully upgrades ESP performance
Finally, for proper operation of a multi-pollutant system containing an ESP, the role of the precipitator is central. Using the non-invasive PALCPE and VESPR techniques can help improve operation of the precipitator, and thereby contribute to the overall performance of the multi-pollutant system.