United States Environmental Protection, Agency Research and Development Industrial Environmental Research = > Laborato o£ Research Triangle Park NC 2711 EPA-600/52-64.019 Mar. 1984 Project Summary Limestone Scrubber Slurry Automatic Control Systems Patrick H. Garrett In the case of environmental processes, consistent reduction in the variance of controlled variables with large through- ut processes is of particular ‘Accordingly, this report utilizes current understanding of limestone scrubbers for fiue ges desulfurization (FGD) to develop an effort into the optimization of automatic control for the recircul slurry processes. To thi acknowledged methods of mathematical ‘modeling, computer simulation, and experimental proofing are applied to the design of slurry limestone addition, ‘slurry density, and absorber (988 ratio control systems. Three automatic control methods are to the recircul experimental control is based on material balance considerations, but is compromised by the lack of slurry reaction measurements. Control of pH is geared toaccommodate variations in the slurry reactions, bt requi high slurry reaction gai {ApH7A limestone fesdrate) for stable {and responsive control. Stoichiometric- assisted pH control offers additional process disorder reduction than either stoichiometric or pH control separately, but ts additional complexity is warranted only under conditions of low si reaction gain, Absorbs ‘energy requirements while maintait ‘an 807 exit target i leo designed based ‘on a feodforward SOz removal law and slurry pump selection. This Project Summary was developed by EPA's Industrial Environmental Re- search Laboratory, Research Triangle Park, NC, to announce key findings of the research project that is fully docu- ‘mented in a separate report of the same title (s00 Project Report ordering information at back). Introduction Early attempts to control Jimestone scrubbers involved combinations of liquid-to-gas ratio, limestone foodrs and recirculating solids which would ‘maximize SO2 removal while minimizing the risks of internal scaling. However, an incomplete understanding of observed effects prevented the achievement of desired goals. More recently the charac- terization and operating experiance with these processes has reached a level of ‘maturity to better support a productive effort into the optimization of their control. The application of automatic control to ‘environmental processes such as lime- stone-alkali wet scrubbers is especially promising because of its potential for increased reliability, economy of opera- tion, and consistent reduction in the variance of controlled variables with varying operating conditions and opera: tor proficiency. Implementation of solu- tions to the major obstacles to limestone scrubbing, the latter including internal scaling and insufficient SOs remov- al, in practice can be achieved only through the disorder reducing capabilities of pro- cess automation. The effectiveness of limestone scrub- bing depends primarily on the ability to ‘accommodate and regulate the manifold factors associated with the recirculating slurry. This problem is all the more @ challenge because of the few usetul slurry process measurements and actua- tor inputs available, and the economic considerations associated with the energy and feedstock requirements of large scale scrubbers. These consequen- cos prompted the use of process modet- ing, computer simulation, and control‘system experimental testing for the development, design, and detailed analy- sis of the recirculating slurry control ‘systems presented in this report. Conclusions and Recommendations ‘Automatic control of the recirculating slurry is essential for successful operation of limestone scrubbers for reliability, SOz removal, and efficiency. Three control methods have been analyzed and specified for the addition of dense limestone to the slurry, which constitutes the most difficult scrubber control action to realize effectively because of competing slurry factors. Stoichiometric control of limestone addition is widespread in application, unconditionally stable, and meaningfully based on slurry and inlet-gas material balance measurements. However, this control method utilizes dead reckoning based on feedforward measurements, and is lacking in its visibility of influential slurry factors which can result in signifi cant errors in the limestone feedrate, Therefore, its utility as the sole determiner of limestone addition is compromised and (therefore) not recommended. Control of the pH of limestone feedrate is geared to accommodate slurry factors which are invisible to stoichiometric control, primarily because slurry pH mea- urement represents an agglomerate of the slurry reactions. However, the imple- mentation of this control method requires 2 rigorous design for stable and responsive performance, since the process intervenes between the controller measurementand limestone feed, Performance is improved with an antiwindup batching controller which prevents further integration of the error signal when the output actuation signal has limit cycled. A nonlinear pH controller is not required for limestone reagent because of the highly buffered dissolution and limited range of neutrali zation reaction pH. A narrow-range 4.5 to 6.5 pH probe was found to increase sensitivity and control responsiveness, and cascade flow control was used to compensate for disturbances in the limestone feed ‘system. Responsive pH contral is achieved for pH set points (<5.8) that result in a high slurry reaction gain (4 pH/ A limestone feedrate). Consequently, this is the recommended limestone feedrate design when conditions of high reaction gain are maintained during scrubber operation. The combination of stoichiometric: assisted pH control utilizes the most independent process measurements to 2 0.40. 038 0.30 ‘Recommended 026 Reaction Gain, AoH/ Agpm Limestone pH Contro)_ >> Stoichiometric Assisted pH Control Recommended 020: 15 0.10 so 52 54 56 88 60 Scrubber inlet pH Figure 1. Reaction gain versus slurry pH. determine the limestone feedrate, and is, ‘optimum in the sense that further reduction in the slurry process disorder is available than either method is capable of individually. Deadband pH and integrat: io control additions necessary 10 realize improved perform- ance result in increased system complex- ity, and is warranted only under the conditions of low slurry reaction gain shown in Figure 1 ‘An analysis of limestone slurry lqui to-gas (L/G) ratio control for minimum scrubber energy requirements, whil simultaneously maintaining an SO2 exit target, resulted in the design of a feedfor- ward control system. This system uses inlet S02 and flue gas volumetric flow measurements with a removal efficiency law and slurry pump selector logic for continuously determining the minimum required L/G ratio. Adding an outlet SOs ‘measurement to compensate the feedfor ward L/G determination for errors in the ‘exit target was found to produce contro! system instability under all conditions and (therefore) is impractical. These results are summarized in Table 1Table. Limestone Scrubber Slurry Automation Research Milestones Element Modeled Sinulted Tested Hightighe PA control of x x 1X. Reconmended contol metod but requires Timestone feodrate his reocton gain (Spas tenestone feocrateh ‘Stoichiometric limestone x x X Rapid conta response to inet changes, leodrte control but errors due o-no Surry messurements ‘Stoichiometric assisted x x X——Meximum slur disorder reduction, but pi comet complony warrented ony for aw reaction oan Absorber quidto-gas x x Energy sovings based on eedtorward control ratio control * using inlet SOs Unstable using outlet SO2 Limestone stry x x Maintains optimum condition for sols density contol ‘natatizaon and scelng prevention ‘Story butter x ato control determined trom percent of ‘ives contol limestone ledrote Patrick H. Garrett is with the University of Cincinnati, Cincinnati, OH 45221. D. Bruce Harris is the EPA Project Otticer (see below). The complete report, entitied “Limestone Scrubber Slurry Automatic Control Systems,” (Order No, PB 84-148 766; Cost: $11.50, subject to change) will be avaifable only from: ‘National Technical Information Service 5285 Port Royal Road ‘Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Industrial Environmental Research Laboratory US. Environmental Protection Agency Research Triangle Park, NC 27711United states Centar for Enwronmental Research Environmental Protection Information Agency Gineinnan OH 45268 Ofticia Business Penalty for Private Use 8300 1 US, GOVERNMENT PRINTING eosene