Integrated Automation System for Rare Earth Countercurrent Extraction Process

JOURNAL OF RARE EARTHSVol .22，No.6, Dec. 2004，p.752 .Integrated Automation System for Rare Earth Countercurrent ExtractionProcessChai Tianyou( 柴天佑)1 , Yang Hui(杨辉)1.2.( 1. Research Center of Automation，Northeastern University, Shenyang 110004，China; 2. Schoolof E-lectronics and Electrical Engineering, East China J iaotong University, Nanchang 330013，China)Abstract: Lower automation level in industrial rare- earth extraction processes results in high productioncost, inconsistent product quality and great consumption of resources in China. An integrated automationsystem for extraction process of rare earth is proposed to realize optimal product indices, such as productpurity ,recycle rate and output. T he optimal control strategy for output component, structure and functionof the two-gradcd integrated automation system composed of the process management grade and thprocess control grade were discussed. This system is successfully applied toa HAB yttrium extraction pro-duction process and was found to provide optimal control, optimal operation, optimal management and re-markable benefits.Key words: countercurrent extraction; intelligent optimal control; integrated automation system; rare earthsCLC number: O614. 33Document code: AArticle ID: 1002 - 0721(2004 )06-0752 - 07.There are abundant rare earth resources in Chi-“Reduction of production cost，improvementna，which has the largest reserves and all kinds of va-of product quality, minimization of environmentalrieties. Rare earth researchers and operators in Chinaimpact and minimization of resources consumptionhave developed advanced rare earth extraction tech-can only be achieved by optimal design of all auto-niques'1] with the largest rare earth and capacity out-matic control facilities” has been pointed out input in the world since 1970' s thanks to the counter-Technology Road Map to Determine the Researchcurrent extraction theory. Efficient and consistent au-Priorities of the European Steel Industry-4]. Thetomatic production processes are in great need for theintegrated automation system of process industriescontinuous, large- scale and integrated rare earth pro-is the key to solve the above problems. The inte-duction. In France, Japan and the U. S. A.，automat-grated automation system is proposed in Ref. [5],ic measurement of material concentration，acidity andwhich consists of five grades: process control,flow rate in rare earth extraction process has been a-process optimization，production scheduling， erchieved as well as automatic control of material feed.terprise management and economic decision. TheAs a result steady product quality is achieved in Ref.control system of minerals processing consisting[2]. Automatic level of domestic equipments are gen-of three grades: process stabilization, process operally low, resting on off-line analysis, manual regu-timization and process management which is de-lation and experienced control that result in infficientveloped in Ref. [6]. The integrated automationproduction，high resource consumption， inconsistentsystem with three grades， Enterprise Resourceproduct qualities, which have become bottleneck ofPlanning ( ERP)，Manufacturing Execution Syrare earth industry. Therefore, it is pointed out intem( MES) and Process Control System (PCS) isRef[3] that the technical target of the tenth “Five-given in Ref. [7]and has been successfully appliedYear Plan”of rare earth production is “the researchto Paishanlou Gold Mine of China. It has beenon new equipment and automatic control technique forfound to provide many benefits to this gold mine.high purity rare earth as well as on-line supervisoryAccording to the characteristics of productionsystem and automatic control technique for rare earthes manaoement of rare earth indus-中国煤化工。extraction".tryautomation system forReeived date: 2004-04- 20;revised date:2004-08 - 10MHCNMHGFoundation item: Project supported by the National Tenth Five Year Plan of Key Technology( 2002BA315A) , the National Key Fundamen-tal Research Program( 2002CB312201) ,and the National Natural Science Foundation of China( 5047 4020).Biography: Chai Tianyou( 1947-emician of Chinese Engineering Academy* Corespgesei易雅uor( E- mail: yhshuo@ 263. net)75ChaiTYet al .Integrated Automation System for Rare Earth Countercurrent Extraction Processextraction process of rare earth by using the struc-stage to extractor at the rate of u1; organic solture of integrated automation system of industrialvent and .organic saponification ammonia are addedprocess is presented in Ref.[8], which is to real-to extractor at the rate of u2 and us respectivelyize the optimal production indices. This systemfrom the first extraction stage according to specif-has been applied to HAB yttrium extraction prod-ic saponification degree; washing liquid enters exuct process and it has been found to provide re-tractor from the last(n+ m)scrubbing stage at themarkable benefits.rate of u4. Extraction stages aim to extract mostProcess of Rare Earth Counter-. A and a little B from aqueous phase solvent toform organic load phase; washing liquid and or-current. ExtractionRare earth consists of seventeen elements, .ganic phase meet at multi-stage in scrubbing stagelike lanthanide, scandium and yttrium，which ex .in order that most B of organic load phase can in-ist commensally in nature. The separation is labo-pour into aqueous phase. The stream of aqueousrious because of the resemblance of chemical prop-phase from right to left and of organic phase in ex-erties of rare earth. The requirement of the puritytractor from left to right is converse because ofof rare earth and other elements are very high inthe special structure of extractor and the forcesome applications, usually at the level of part perproduced during stirring. Aqueous productionmillion. So higher demand on the rare earth ex-contains B whose purify is y1 from the first extraction is presented.traction stage after exchange and purification be-The liquor of mixed rare earth is separated, en-tween each extraction stage and washing one，asriched and extracted to achieve agreed purity and recy-well as organic production from scrubbing stage ncle rate in rare earth countercurrent extraction. Real十m containing A whose purify is y2 are obtained.process of rare earth extraction separates lanthanumMultistaged( usually from decades to hundreds)(La)，cerium(Ce)，praseodymium(Pr)，neodymiumseparation process are characterized by complex mech-(Nd)etc. The separation point is fixed between Laanism and complex extraction equilibrium reactionand Ce,Pr,Nd... ,therefore La is marked B as ele-whose mixed time is about four to seven minutes andments extracted hardly with others as A extractedsettlement time is fifeen to twenty one minutes ineasily, while fixed between Ce and Pr,N.... ,La,Ceeach stage due to varieties of rare earth more than tenwill be regarded as B and Pr ,Nd.... integrated as A.compositions，intense change of component， multi-Meanwhile A/B is denoted. Fig. 1 shows the flowfactors affecting separation efficiency and small sepa-chart of rare earth extraction. From left on, it is ex-ration coefficients. Regulating control variables liktraction process composed of n mixer settlers stagesflow of extraction liquid, washing liquid or feed usu-and washing process composed of m mixer- settlersally affect purify of outlet production long after severalstages.hours' even several days' transfer. Therefore, processIn Fig. 1, uu denotes the flow rate of feed solvent,detecting points are set between five and fifteen stagesu: the flow rate of organic solvent,us the flow rate offrom outlet through measuring and controlling com-organic saponification ammonia, us the flow rate ofponent content(y3 ,ys )to achieve high purify(y1 .y2).washing liquid in the aqueous phase; y1 denotes the pu-However, component content of rare earth is hard tority of production from aqueous phase outlet,yz the pumeasure online and the countercurrent extraction hasrity of production from organic phase outlet, Ys and yicomplexities like multivariable， strong decoupling,the content of rare earth component on detectingdelay and time variance,and it is difficult to get opti-points.mal control.The typical extraction process includes raw mate-rial disposal, countercurrent extraction, dehydration,Serub.roasting，mixing and packing，so whatever fault of a-....n+1| ..-.n+mbove mentioned part will take effect to the wholeproc中国煤化工ary to make reasonable↓y↓y,prMHC N M H Gre scheduling for eachFig.1 Description of rare- earth countercurrent extrac-part in order that the process can realize the resourcestionmanagement and effective scheduling in special status.Aqueous feed solvent including A and B areThe scheduling of key technics variables is significant ,expected两方数据parated flowing from n( feed-in)for it can assure the quality and output of production.754JOURNAL OF RARE EARTHS, Vol .22, No.6，Dec. 2004Efficient and continuous operation of resources as well2. 1 System structureas linking and parallel of working procedure are nec-Adopting the system structure of integratedessary in countercurrent extraction process， amongautomation system for industrial process in Ref.which there are problems of resources balance and[8]，it gives the integrated automation system fortime balance. Resources balance is to realize the corrare earth countercurrent extraction by combiningrespondings of production planning and productionwith process characteristics in Fig. 2. This systembatch. Time balance such as establishment of produc-realizes optimal production indices like purity, re-tion line train schedule is to realize utilizing rate andcycle rate and output.eliminate production bottleneck. The planning man-This system consists of process control sys-agement of working procedures can realize balancedtem, production management system and comput-load, close linking of resources and to ensure flexibil -er support system. Process control system of ex-ity，quality and delivery time. Otherwise， it cantraction adopts integrated design technology andchoose optimal resources path to decrease cost，energyadvanced control technology of EIC( Electric Iresources cost and other ullage then to make less pro-strument Computer)distributed PC control systemduction cost and more economy benefits. At present, .and object oriented intelligent optimal controlalmost enterprises of countercurrent extraction adopttechnology. The system consists of basic automa-manual production planning and production schedulingtion technology， optimal control technology anwith low efficiency, unreasonable plan and unpunctualmultimedia monitoring system. Production man-planning adjusting. As a result, it brings low efficien-agement system adopts case-based productioncy of production, high cost and worse competition a-planning scheduling technology in Ref. [9]，dy-bility.namic cost control technology of production2Integrated Automation Systemprocess in Ref.[10] and object oriented optimalfor Rare Earth Countercurrentoperation and optimal management technologies. .It consists the fol-Extraction Processplan managementproduetionpnrduetionpquipmentmanagementpedurtin srie |enr plamnamning a | protuctsystemplanPandurtion material| energy004t eontmlStariaiseamanare manags-rquipnuent quality| view ing aundSrherduting| nentmenmanagrmentmanageroent contraLmmpport .turnageme ol pocluetion proucsinformation proesing system nf prudurstion processeathContrnltntoptimal ceontrd Kystetm forsuhsystemsextrarlinn pruresmutimedlia|l dataclepsition annd.rouating rontrolmonitoringionp contxuhy slems Fordiaposl-forwant anddisposalbark rontroluhsystemkprucess contfulFystremmurlting stnrk, hatclh, material transponruation, rxirartion F二Rare earthraw oresC，ruetionFig.2 Integrated automation system for rare earth countercurrent extraction processlowing subsystems: production planning andof production management system and processscheduling, data statistics and analysis, cost con-control system by information disposal system. Sctrol and management, equipment operation man-th中国煤化工process managementagement，production quality management， inte-ar:MYHCNMHGiswellastheoptimalgrated query and assistant decision support sys-control，optimal operation and optimal manage-tem. Consisting of the relational database， real-ment of rare earth extraction. .time database and computer network system, the2.2 System functionscomputesystem can make the integrationFig. 3 shows the structures of system func-75ChaiTYet al .Integrated Automation System for Rare Earth Countercurrent Extraction Processtions of production process management, whichduction plan. The subsystems can control the ma-consists the following subsystems: production op-terial cost and energy resources by analyzing theeration management，material management, costdata according to the material cost plan and energymanagement, energy management，quality man-resources cost plan. They can also measure theagement，equipment management，production in-material cost and energy resources cost on line andformation query and assistant decision supportpredict the cost， then give the operation instructionsystem.on how to decrease the dynamic cost according to theBased on the integrated production indicesproduction plan. Monitoring the equipment operationlike output，purity and recycle rate，productionstatus in real time, they can predict the operation state,process management system can make the plans ofset the dynamic maintenance management plan to insureproduction，material cost，production cost, enethe safe operation. They can realize the real time monigy sources cost， quality， operation and mainte-toring and management of production quality, optimalnance equipment for the preprocessing， counter-operation condition by collecting and analyzing the pro-current extraction and postprocessing. Productionduction quality data of raw material , semi- manufacturedmanagement subsystems transform those indicesgoods and finished product to guarantee the optimal op-into the output purity indices by technics simula-eration when the condition and flow are changed.tor，then dispatch the production process in realFig. 4 shows the function structure of the controltime, analyse the statistical data to insure the in-system. It is consisted of basic automation system, optegrated production indices according to the pro-timal control system and multimedia monitor system.Productionmanagement systemsimulatoranalysis of integrated production indices[ planning and schedulingOperationmanagement[ data stistisC data analysisL material planMaterial[ matenal providing andulatingmanaemermaterial balance calculatingC cost plandata analysis of material depletionCost厂cost monito[ cost Tckoningcost analysis and contolL energy planEnergyC energy data acquisitionenergy consuming analysisC quality indexQuality[ quality stisticsquality ananysisquality monitor and control[ equi pment accountEquipmentL equi pment maintenance plan[ operation monitorng.controlplan scheduling information[ equi pment fault diagnosisInfomationproduct quality informationmateraldepletioninfommationinquiry andnationdecisionenergy depletion informationequipment status informationCcost informationassistant decision supportFig.3 Functions of production management system for rare earth extractionbatch controlDisposal-foward= Tou= suhsystemtransportation controlBasic automation[ led-in Aux controlf Disposal-back suhsystematerial iquid levelMultimedia.Data acquisition sytlemcontmolmonitoringReal time monilnring forexlraction mixing syslrimReal time monitoring Tor1|_ concentration andExtraction,L extraction中国煤化工optimal controlTealtime1HC NMH G.record,.nLe:-.for key informalionFig.4 Functions of process control system for rare earth extractionTh男数据tomation system consists of thepre- processing control subsystem, process control756JOURNAL oF RARE EARTHS, Vol .22, No.6，Dec. 2004subsystem，post processing control subsystemis proposed, which is the combinations of of-lineand measure data acquisition system， these sys-assay test and component content soft sensor, oftems can realize the logic control of equipment andtimal setting and loop control, process modelingthe loop control of material level meter and flowand control, the optimal strategy in Ref. [11] andrate as well as the real time data acquisition of ma-characteristics of extraction process. The strategyterial and energy sources cost. The basic automa-can achieve the optimal setting of flow rate pointstion system and multimedia monitoring system areand revision of feedforward and feedback to getunited to realize the real time monitor of spooptimal control of output purity indices.scene, operation state of equipment，key parame-By objective transformation model of optimalters and trend curve so that the stable operationcontrol system, the A output purity and B outputcan be gained. Otherwise, the multimedia monito-purity indices can be transformed into indices ofring system can scout, record， expand, monitorcomponent content. Case- based intelligent pre-in a long- distance and diagnosis fault for the infor -setting model02J brings the optimal set points ofmation of key posts and equipments. Extractioncontrol loop and feed flow rate of extractionoptimal control system can control the content ofprocess can follow the optimal set points by looprare earth component at detecting points accordingfeedback control. Input and output detectingto optimal setting and feedback revision for mate-points and soft sensor model [18] of componentrial flow rate by the indices of export output puri-content are utilized to give the predict points,ty，then the output purity indices can realize opti-which are compared with the ideal indices. There-mal control.fore, the errors can revise the optimal set points2. 3Optimal control strategy for outputof control loop for flow rate by feedforward com-puritypensation. The assay points are compared with i-According to complicated industrial process,deal indices， errors can revise the soft sensorsuch as rare earth countercurrent extraction,model of component content by adaptive mecha-which has the characteristics of multivariable,nism to improve the predict precision and to real-strong decoupling, large delay, nonlinearity, timeize optimal control of extraction, which is thoughvariance as well as the difficult on-line measureforset points of feedback revision control loop, assaycomponent content, it is difficult to control withpoints of component content and predict points ofconventional optimal control methods. So， the in-soft sensor model.telligent optimal control strategy shown in Fig. 5nTtimecomponent|nT time「Fedback+ 当一 asay value fronopmenconten| purifyFeed7 optmalC compensationdeterting.L luxsel pointsTorpeints。置县Organic. lor Lpts,tractiondertionHBpurify、自Trimeadaptiverevisininstitutionfeedfrwardlompensation+[ samogeasurementTtime +↓- predict value foromponent contentFig. 5 Intelligent optimal control for rare earth countercurrent extraction process3 Application in Yttrium Extractionting stages. The first stage contains sixty extrac-Process by HABtors requiring to achieve Y2O3 whose purity isA company extracted high purity yttriummore than 99 % and provides materials for the sec-from ionic rare earth, in which the content ofondstaoe: the serond staoe is composed of eightyY2O3 is more than 40%，adopting new extractionext中国煤化工，ith purity more thantechnique of HAB dual solution. This technique99%:MYHCN M H Gts of forty extractorsrequires the purity of ultimate yttria to be moreto extract impurity from high purity Y2O3. In thethan 99. 99% and of mixed rare earth containingprocess of HAB extracting yttrium, control of thelow purity yttrium to be less than 0. 5%. Thefirst stage is the key for the implementation of o-whole extht 6括process consists of three extrac-verall production index of extraction process. .75ChaiTY et al.Integrated Automation System for Rare Earth Countercurrent Extraction Process3.1 Design for system hardwareProcess control ofcountereurrent extractionAccording to the characteristics of HAB ex-tracting yttrium and structure system of integrat-厂OpimaT controlo[procesr monitoring oftounterrurrent exiratbunterurrenl extractioned automation system for extraction shown in Fig.2，the integrated automation system of HAB ex-tracting yttrium is built. Its hardware structure is到shown in Fig. 6. One model computer, one set ofoperator monitoring station and one multimediamonitoring system，three groups of system con-trol cabinet consisted of PLC and frequency con-verter are equipped in central control room. ModelFig. 7 Software structure of extraction process controlcomputer is to achieve soft sensor and optimaltem. The combination of all the interfaces carcontrol. Operator monitoring station is used forgive the operation status and real time changes ofon- line parameters monitoring and soft manual op-key parameters clearly.Control software oeration. Multimedia monitoring system is to mo-process uses Step 7 of Siemens，then sequentialnitor the key equipments and posts. Profibus net-control, loop control and liquid level control are a-work is adopted between PLC and frequency con-chieved. Multimedia monitoring software adoptsverter，Ethernet is adopted between computer andVC++,so the monitoring， recording， replayingPLC.and long distance monitoring of key equipment,post and images are obtained.广Operatorf Modeling Multimedia 下work station work slalionworkstationyThe optimal software of HAB extracting yt-( Ehemettrium consists of soft sensor software and threeseries optimal software. Refer to Fig. 5，the opti-I Profibusmal control are implemented as follow:L netwol(1) Enter production purity indexof A and Band border production conditions， such as feedconcentration, rare earth distribution, extraction面d-由一面”solvent concentration and scrubbing liquid acidity,圓etc.(2) The pre-set points of control loop can beFig. 6 Hardware structure of integrated automation system inachieved by case based intelligent optimal settingyttrium extraction by HABcontrol technology.Hardware of production management systemincludes PC and server. Database is consisted of(3) Enter the predict points into soft sensorreal time database and relation database and canmodel, the prediction points of every stage andintegrate information of production managementcomponent content can be obtained， which are :system and process control system by control net-compared with target points after optimal settingwork( Profibus DP) and Ethernet. The operatorsin the range of target points and can load the pre-can realize on-line management by managementdict points of every flow rate. Then turn to thesystem and real time control by process controlfifth step.system. .(4) If prediction points of the soft sensor3.2 Design for system softwaremodel are not in the range of target points, revi-Control system software of HAB extractingsion module will be started to adjust the predictyttrium is consisted of monitoring software andpoints of flux. Then turn to the third step.optimal control software, which is shown in Fig.(5) When optimal process is finished， the7.system will deposit relevant information automati-Monitoring system of HAB extracting yttri-call中国煤化工on-line revision andum adopts WinCC of Siemens, consisting of staticofMYHCNMHGinterface and dynamic interface. The static inter-3.3 Application resultsface reflects the configuration status of main e-This system was put into trial run in Septem-quipments and the dynamic interface reflects theber, 2003. Fig. 8 gives yttrium product purityreal time bp教循n status of equipments and sys-curve of the first section at the first stage. Fig. 9758JOURNAL oF RARE EARTHS，Vol .22, No.6，Dec. 2004gives non- yttrium product purity curve of thetraction successfully, the optimal control, optimalfirst section at fifty-two stage of HAB yttrium ex-operation，optimal management and remarkabletracting. The integrated automation system carbenefits have been achieved. This system can alsogreatly decrease the need for assay of yttrium.be applied to other hydrometallurgical processesThe content of yttria impurity in non- yttriumbesides rare earth countercurrent extraction.phase is less than 0. 5%,the yttria purity is moreReferences:than 99. 5%。Recycle rate of yttrium rises by 2%，[1] Xu Guangxian. Rare Earths( in Chin. )[M]. Beijing:which assures continuous and stable production ofMetallurgical Industry Press, 1995.[2] Yan Chunhua， Jia Jiangtao, Liao Chunsheng, et al.high purity and recycle rate and achieves remarka-Automation system in rare earth separation by coun-ble benefits.tercurrent extraction process[J]. Chinese Rare Earthsin Chin(in Chin. ),1997，18(2): 36.[3] National Development and Reform Commission. Rare99earth industry tenth five- year development programsin China(in Chinese)[P],2001.8.5-[4] Charbonnier et al. Technology Road Map to Deter-r 985mine the Research Priorities of the European Steel In-.97.5dustry[R]. 1999.[5] CIM Reference model committee, Purdue University.97 tA reference model of computer integrated manufactur-96.5-ing from the viewpoint of industrial automation [J].96Int. J. Computer Integrated Manufacturing, 1989, 2(2):114.9510Date2[6] Houseman L. A, Schubert J. H. ,Hart J. R. A plant-wide control platform for minerals processing [J_Minerals Engineering, 2001 ,14(6):593.Fig.8 Purity curve of yttrium at the first stage[7] Chai Tianyou, Li Xiaoping, Zhou Xiaojie, et al. Con-4 Conclusiontemporary integrated manufacturing system based onObject -oriented integrated automation systemERP/MES/PCS in gold mine enterprise [J ]. Controlis proposed, which adopts system structure of in-Engineering of China(in Chin. ), 2003, 10(1):18.[8] Tianyou Chai. Integrated automation system for min-tegrated automation in Ref. [8] according to char-erals processing [A]. The Plenary Lecture of the IF-acteristics of countercurrent extraction realize au-AC Workshop on New Technologies for Automationtomation for rare earth countercurrent extractionof Metallurgical Industry[C], Shanghai , China, Oc-process. This system consists of production man-tober 2003.[9] Zheng Binglin, Hu Kunyuan, Chang Chunguang. Sta-agement system and process control system. Pro-tus and expectation of research on integrated planningduction management subsystems transform inte-for steel production [J]. Control Engineering of Chinagrated production indices into the output purity(in Chin. )，2003,10<(1):6.indices, this strategy can realize optimal control.[10] Li Xiaonan, Chai Tianyou. Dynamic cost control meth-od in production process and its application[ A].15thThis system has been applied to HAB yttrium ex-IFAC World Congress[C]. Barcelona Span. July 21 - -00厂26 , Elsevier Science Ltd, 2002.9.5中[11] TianyouChai， shoupingGuan. Object-Oriented Integrated Control Technology of Complex Industrial8.stProcesses[A]. The Plenary Lecture of Proceedingsof IFAC 5th Symposium on Low Cost Automation9[C]，Shenyang, China, 1998.297.5[12] Yang Hui, Tianyou Chai. Optimal supervisor control97of rare earth extraction separation processes based on96.5case reasoning [J]. Journal of Northeastern Universi-ty( Natural Science).[13] Yang Hui, Tan Minghao, Tianyou Chai. Neural net-95.5_works based component content soft sensor in coun-9S中国煤化工rion [J]. Journal of RareMHCNMHGFig.9 Purity curve of non-yttrium at the fifty-two stage.