Board of Research in Nuclear Sciences

GOVERNMENT OF INDIA

DEPARTMENT OF ATOMIC ENERGY (DAE)

BOARD OF RESEARCH IN NUCLEAR SCIENCES

Project Proposal Application Form

BRNS-2A

SECTION-A

PART I – PROJECT OVERVIEW

…

100. Advisory Committee Code Number: 34

101. Title: Dynamic Data Driven Model Based Predictive Control of Nuclear Steam Generator

102. Key Words & Name of 3 Referees:

Steam Generator Control, Shrink and Swell Phenomenon, Time Series Modeling, Multi-variable Control, On-line Constraint Control, Model Predictive Control, Adaptive Control, Embedded Control Systems

103. Project Summary:

The water level of a nuclear steam generator (SG) must be properly controlled to secure sufficient cooling water for removing the primary heat and to prevent the damage of turbine blades. The inadequate performance of the conventional controllers for the SG water level can result in reactor shutdown.Recently, Model Predictive Control (MPC) technology has been employed in French nuclear power plants for better level control in SG and reducing down-time. Ability to handle multivariable interactions and operating constraints systematically is the key feature of MPC technology. It has been widely used in process industry over last three decades for controlling many complex unit operations. This project aims at the development of control relevant dynamic modeling and MPC software for controlling nonlinear and time varying level dynamics in SG. This objective will be achieved by developing suitable methods and tools for (a) Data Driven Modeling: Development of tools for off-line and on-line dynamic time series models suitable for model predictive control (MPC), (b) Constrained Nonlinear Programming for solving constrained nonlinear optimization problems online and (c) Model Predictive Control capable of robustly controlling moderate scale multi-variable process. The proposed development will be carried out using numerical simulation studies and experimental evaluations on (i) Numerical simulation studies using detailed dynamic models of steam generator developed using first principles in collaboration with PC (ii) Experimental studies at IIT Bombay using experimental rigs available at Automation laboratory, IIT Bombay using embedded control hardware (iii) Experimental evaluation at steam generator simulation facility at BARC. Systematic evaluation of possible benefits that can be accrued by implementation of MPC schemes in terms of operating efficiency, improved handling of load disturbances and transitions in operating points, reduction in operating costs and energy costs will be carried out and benchmarked against conventional control approach.

104. PI's Name & Address105. CI's Name & Address 106. PC's Name & Address

Prof. Sachin C. Patwardhan
Dept. of Chemical Engineering
I.I.T. Bombay, Powai
Mumbai 400076 / Prof. Santanu Bandyopadhyay
Energy Systems Engineering,
I.I.T. Bombay, Powai
Mumbai 400076 / Dr. C. Vijayan
B. A. R. C., Mumbai
Dr. Kallol Roy,
B. A. R. C., Mumbai

110. Budget Amounts in Rupees (Please see Instructions - Sr. No. 14 to 20):

Particulars / Year I / Year II / Year III / Total
310 / Equipment: / 1200000 / 0 / 0 / 1200000
320 / Staff Salaries / RA I: / 171600 / 171600 / 171600 / 514800
RA III: / 280800 / 280800 / 280800 / 842400
330 / Technical Assistance: / 60000 / 60000 / 60000 / 180000
340 / Consumables: / 25000 / 50000 / 50000 / 125000
350 / Travel / PI: / 40000 / 30000 / 30000 / 100000
PC/DC: / 10000 / 20000 / 20000 / 50000
360 / Contingencies: / 40000 / 30000 / 30000 / 100000
370 / Overheads: / 274110 / 96360 / 36780 / 407250
380 / Total / 2101510 / 738760 / 679180 / 3519450

111. Equipment and Cost (Please provide here the name of the equipment and cost only):

Item / Year I / Year II / Year III / Total
1 / Simulation Workstations and Printers (2 Nos.) / 150000 / 0 / 0 / 150000
2 / Industrial Grade PCs for On-line Control (2 Nos.) / 150000 / 0 / 0 / 150000
3 / On-line Data Acquisition and Embedded Control Hardware / 400000 / 0 / 0 / 400000
4 / Software Libraries for optimization and control / 300000 / 0 / 0 / 300000
5 / Hardware modifications in distillation column setup / 200000 / 0 / 0 / 200000
Total / 1200000 / 0 / 0 / 1200000

PART II - PROJECT OBJECTIVES, RESEARCH PLAN and DELIVERABLES

...

200. List of Objectives (Please see Instruction - Sr. No. 12):

Overall objective: Development of control relevant dynamic modeling and online model based predictive control schemes for multivariable constrained control steam generator system in a nuclear power plant.

Development of Tools: This objective will be achieved by developing suitable methods and tools for

Data Driven Modeling: Development of tools for off-line and on-line dynamic time series models suitable for model predictive control (MPC).
Constrained Nonlinear Programming: Efficient sequential linear programming (SLP) and sequential quadratic programming (SQP) for solving constrained nonlinear optimization problems online. Efforts will be made to develop libraries suitable for embedded control implementation.
Model Predictive Control: Capable of robustly controlling moderate scale multi-variable (roughly 10 inputs, 10 outputs) systems

Simulation and Experimental Studies: The above development will be carried out using numerical simulation studies and experimental evaluations on

Numerical simulation studies using detailed dynamic models of steam generator developed using first principles in collaboration with PC
Experimental studies at IIT Bombay using (a) Heater-Mixer setup (b) Packed-bed distillation column setup available at Automation laboratory, IIT Bombay
Experimental evaluation at steam generator simulation facility at BARC.

Cost Benefit Evaluation: Systematic evaluation of possible benefits that can be accrued by implementation of MPC schemes in terms of operating efficiency, improved handling of load disturbances and transitions in operating points, reduction in operating costs and energy costs and benchmarking with existing control scheme.

210. Yearly ResearchPlan and identify the Deliverables (Please see Instruction - Sr. No. 13):

A. At PI/ CI's Institution

Ist Year:

S. No. / Activity / Quarter 1 / Quarter 2 / Quarter 3 / Quarter 4
1 / Development of SG Simulator / / /
2 / Control Relevant Modeling Tools / / /
3 / Procurement of Hardware and Software / /
4 / Installation and testing of hardware / /

IInd Year:

S. No. / Activity / Quarter 1 / Quarter 2 / Quarter 3 / Quarter 4
5 / Control Relevant Modeling of SG Simulator / /
6 / Optimization Tools: Development & testing / /
7 / MPC of SG using Numerical simulation / /
8 / Cost –benefit Analysis / /

IIIrd Year:

S. No. / Activity / Quarter 1 / Quarter 2 / Quarter 3 / Quarter 4
9 / Testing of Embedded MPC using Heater-Mixer Setup /
10 / Testing of Embedded MPC using Distillation column /
11 / Dynamic Modeling using data available from BARC / / /
12 / Preparation of Report /

B. At PC's Institution.

Ist Year:

S. No. / Activity / Quarter 1 / Quarter 2 / Quarter 3 / Quarter 4
1 / Assistance in Development of SG Simulator / / /
2 / Procurement of Hardware and Software / /
3 / Installation and testing of hardware / /

IInd Year:

S. No. / Activity / Quarter 1 / Quarter 2 / Quarter 3 / Quarter 4
5 / Assessment of SG Simulator / /
6 / Data generation for Control Relevant Modeling from BARC Simulator / /

IIIrd Year:

S. No. / Activity / Quarter 1 / Quarter 2 / Quarter 3 / Quarter 4
9 / Evaluation of Control Relevant Models /
10 / Possible Implementation of MPC on a small sub-section of BARC Simulator / /
11 / Preparation of Report /

PART III - BUDGET ESTIMATES

…

300. Details of budget requirements (Please see Instruction - Sr. No. 14)

BUDGET DETAILS

…

310. Details of the budget for equipment to be procured by the PI (Amount in Rupees):

Local:

Imported:- Nil -

------

Total 1200000 1200000

------

320. Details of staff salary (Amount in Rupees):

Staff / Nos. / Monthly Salary / I’st Year / II’nd Year / III’rd Year / Total
RA I: / 1 / 11000+HRA / 171600 / 171600 / 171600 / 514800
RA III: / 1 / 18000+HRA / 280800 / 280800 / 280800 / 842400
Total / 2 / 29000+HRA / 452400 / 452400 / 452400 / 1357200

340. Details of budget for consumables to be procured by the PI(Amount in Rupees):

Consumables / Year I / Year II / Year III / Total
Printer Cartriges / 12500 / 12500 / 12500 / 45000
Computer Stationary, Memory and Storage devices / 12500 / 12500 / 12500 / 30000
Chemicals for operating distillation column / 0 / 25000 / 25000 / 50000
Total / 25000 / 50000 / 50000 / 125000

350. Details of travel (Amount in Rupees):

Year I / Year II / Year III / Total
PI: / 40000 / 30000 / 30000 / 100000
PC/DC: / 10000 / 20000 / 20000 / 50000

351. Proposed number of visits

of PC/DC to PI's institute

352. Duration of stay during each visit

(No. of days)

353. Total funds required

354. Proposed number of visits

of PI to PC/DC's institute

355. Duration of stay during each visit

(No. of days)

356. Total funds required

357. Funds required by PI for travel to attend conferences within India.

Year I / Year II / Year III / Total
PI: / - / 10000 / 10000 / 20000
PC/DC: / - / 10000 / 10000 / 20000

358. Funds for Other visits (please give details)

Year I / Year II / Year III / Total
PI: / 40000 / 20000 / 20000 / 80000
PC/DC: / 10000 / 10000 / 10000 / 30000

The above funds are reserved for (a) attending annual progress review meetings (b) visiting DEA sites where SG simulation facility is available
BUDGET JUSTIFICATIONS

…

310. Equipment:

Item / Justification
1 / Simulation Workstations and Printers (2 Nos.) / Development of detailed dynamic simulation programs for steam generator and associated systems system using mechanistic models
2 / Industrial Grade PCs for On-line Control (2 Nos.) / Implementing PC based and embedded control on heater-mixer setup and packed bed distillation column setup
3 / On-line Data Acquisition and Embedded Control Hardware / Fast dedicated microcontroller platform that can help decouple control law and associated on-line modeling computations from secondary activities as data display and storage
4 / Software Libraries for optimization and control / Stand-alone C++ program libraries for constrained nonlinear programming that can be used for modeling and on-line predictive control implementation in real time
5 / Hardware modifications in distillation column setup / Modifications in control panels and purchase accessories for being able to control distillation column setup using embedded hardware

320. Staff:

Staff / Qualification / Justification
RA I (Research Engineer) / M. Tech. or B. Tech. with experience / For working on dynamic modeling and steam generator and performing simulation / experimental studies
Expected to register for Ph.D.
RA III (Post Doctoral Research Engineer) / Ph.D.
or research scholar student who has submitted thesis / Post-doctoral researcher versed in advanced control and system identification for developing control relevant modeling and predictive control software

330. Technical assistance:

Employing additional engineer for a short term (say 3 or 6 months) to carry out project related specific tasks (such as programming, interfacing) either full time or part-time employment to post graduate student at IIT for performing similar task.

340. Consumables:

Printing of research papers, reports, patent documents related to project available from the net

Printing of program listings, intermediate simulation and experimental results while carrying out development

Periodic storage of data and software generated as a part of the project

350. Travel:

Visits by PI / PC to DAE site in Rajasthan where dynamic simulator for SG is available

Visits by project staff to DAE site in Rajasthan where dynamic simulator for SG is available for carrying out simulation studies

Presenting parts of research work carried out as part of the project in national conferences.

Visits by PI / PC to attend annual progress review meetings

360. Contingencies:

Advertisement for staff selection

Purchase of books related to project work

Purchase of electronic components for computer and process interfacing

Labor charges for carrying out modifications in existing experimental setups for conducting experiments related to the project

PART IV - OTHER PROJECTS

...

410. List all previous projects that are supported by BRNS or any other funding agency in which PI is actively participating (either as PI or as CI):

------

Sl. No.Title of the projectTotal costAgencyPresent status

------

---N. A.---

------

411. List all projects submitted during the current financial year by PI to BRNS or any other agency for funding. Give details on the present status of the application:

------

Sl. No.Title of the projectTotal costAgencyPresent status

------

---N. A.---

------

412. Brief description of the project(s) submitted/sanctioned by/to PI by other agencies. (Please see Instruction - Sr. No.21):

No on-going project by funding with any government agency or with industry.

Sponsored Research Projects

Sponsoring Agency / Title of Project / Cost(in Lakh Rs.) / Period / Status
1 / A.I.C.T.E.,
New Delhi
(PI) / Model Predictive Control of Nonlinear Processes Using Laguerre Models / 7.0 / April, 96 to Oct., 98 / Completed
2 / D.L.R., Germany (Indo-German Research Program)
(CI) / MERMAID – Model Based Environmental Resource Management Aid / Travel grant / Jan., 96 to
Dec., 99 / Completed
3 / D.S.T., New Delhi
(PI) / Advanced Control and Intelligent Monitoring using a Networked Controller / 15.0 / Feb.,2001-
Dec. 2004 / Completed
4 / M.H.R.D., New Delhi
(PI) / Computer based Automation of Experimental Rigs used in Undergraduate Process Control Laboratory Course / 16.0 / March, 03
March, 05 / Completed
5 / B.R.N.S., Mumbai
(CI) / Development of on-line fault detection & diagnosis (FDD) methodologies in integrated scale complex plants / 26.0 / April, 02-July, 06 / Completed

Industrial Consultancy Projects

Organization / Title of Project / Amount Lakhs Rs. / Period / Status
1 / Infranet System Integrators, Chennai and TLON GmbH, Germany
(Principal Investigator) / Development of a Hardware Module using Multi-Chip Module (MCM) from TLON GmbH, Germany / 1.5 / March, 98 to
March, 2000 / Completed
2 / Honeywell Technology Solutions Lab., Bangalore
(Principal Investigator) / Advanced Planning and Scheduling for Paper, Pulp and Printing Industries (*) / 11.2 / March, 03 – Feb., 04 / Completed
3 / Honeywell Technology Solutions Lab., Bangalore
(Principal Investigator) / Development Of Intelligent Wireless Control Networks
Phase I: Wireless Control of Distillation Column / 6.09 / May,04 – July,06 / Completed
4 / 4. Honeywell Technology Solutions Lab., Bangalore
(Co-Investigator) / Model Maintenance in Advanced Process Control Schemes / 5.0 / May,04 – July,06 / Completed

420. List all previous projects that are supported by BRNS or any other funding agency in which CI is actively participating (either as PI or as CI):

------

Sl. No.Title of the projectTotal costAgencyPresent status

------

---N. A.---

------

421. List all projects submitted during the current financial year by CI to BRNS or any other agency for funding. Give details on the present status of the application:

------

Sl. No.Title of the projectTotal costAgencyPresent status

------

---N. A.---

------

422. Brief description of the project(s) submitted/sanctioned by/to CI by other agencies. (Please see Instruction - Sr.No.21):

No on-going project by funding with any government agency or with industry.

Sponsored Research Projects

Sponsoring Agency / Title of Project / Cost(in Lakh Rs.) / Period / Status
1 / M.H.R.D., New Delhi
(PI) / Modernisation of Energy Systems Laboratory / 20.0 / February, 04
February, 06 / Completed

Consultancy Projects

Organization / Title of Project / Amount Lakhs Rs. / Period / Status
1 / Heavy Water Board
(Co-Principal Investigator) / Development of Helium Production Technology / 5.51 / February 2006 / Completed
2 / Ingenero Technologies (India) Pvt. Ltd.
(Principal Investigator) / Expert Advice for Pinch Study / 0.78 / November 2006 / Completed

SECTION-B(Please submit Fifteen Copies stapled to Section-A)

500. TECHNICAL INFORMATION OF THE PROJECT IN DETAIL(Please see Instruction - Sr. No. 22): In-depth details should be provided in this section on as many number of sheets as deemed appropriate. This information will help us to comprehensively review the proposal.

Detailed Description of Methodology

Need for Efficient Control of Nuclear Steam Generator (SG)

The SG is a highly complex, nonlinear, and time varying system. Its parameters change largely according to changes in operating conditions. The water level of a nuclear steam generator (SG) must be properly controlled in order to secure sufficient cooling water for removing the primary heat and to prevent the damage of turbine blades. Poor control of water level in secondary circuit of nuclear power plant can lead to frequent power shutdowns. Violations of safety limits on water levels are common at low operating power. It may be noted that 30 % of emergency shutdowns of French nuclear reactors have been attributed to problems with level control steam generators. Main difficulties associated with steam generator level control can be summarized as follows (Kothare et al, 2000)

Steam generator level exhibits swell and shrink phenomenon (inverse response): systems with inverse response are inherently difficult to control using conventional PID controllers.
Flow measurements are unreliable at low powers and cannot be used for feed-forward control
System dynamics is highly nonlinear and a simple PID controllers are unable to achieve good control of such systems
Feed water system can only deliver a limited throughput of water to steam generator. Similarly, there are operating constraints on level in steam generator. Conventional PID controllers cannot deal systematically with such constraints, which can lead to poor closed loop performance.

The inadequate and insufficient performance of the conventional controllers for the SG water level has often resulted in reactor trip. Thus, in order to achieve better control of steam generator, there is a need to use control scheme that can deal with process non-linearity, inverse responding behavior, multi-variable interactions and operating constraints in a coordinated manner.

When the SG of the nuclear power plant is operated at low power, the water level can be significantly perturbed. This is the major factor causing reactor trip during the start-up operation. In SG water level is contribute by the mass capacity, reverse dynamics, and mechanical oscillations. The low water level should be prevented so that enough water is supplied to the secondary side. Therefore, the control of the SG water level is important to determine power plant behavior with changes in the operating load condition. It has critical effects on the reactivity control, pressure control of the reactor, and the structural stability of blades turbine blades.

The shrink and swell phenomena in SG are observed in water level responses, when changes occur in the steam flow rates and in the feed water flow rates. The dynamic response of the SG water level is a function of the feed water temperature due to change in feed water flow rates. If feed water flow rate is increased then it will adds mass to the SG, it was expected to increase the measured down-comer water level, it increases only at high power. But at low power, the addition of the cold feed water will decrease in the vapor content of tube bundle, shift in the liquid from down-comer to tube bundle, and temporary decrease in water level this is known as shrink. Similarly, a decrease in feed water flow can cause a temporary increase in water level, which is known as swell. These phenomena are caused by variations in the pressure due to change in the feed water and steam flow rates. These reverse dynamics are short-term effects because the propagation of the pressure is very short. Since SG possesses the non-minimum phase property, the control loop has to be designed carefully to avoid onset of instability. The major difficulties in designing control system for the SG arise from a number of factors (Kothare et. al., 2000)

Nonlinear plant characteristics: The SG is highly complex, non-linear, and time-varying system. Particularly, its parameter undergoes large changes according to changes in operating conditions.

Non-minimum-phase plant characteristics: The plant exhibits strong inverse response behavior, particularly at low operating power due to the so-called “swell and shrink” effects. This non-minimum phase characteristic limits the achievable controller bandwidth.