Preliminary Passive Feedback Model Development and Integration

Abstract

In this report, we document the investigations of feedback models for the passive control systems being designed as a part of this NEUP project. This work builds on our previous efforts that performed Thermal Hydraulics (TH) analysis of the Holos microreactor with the Systems Analysis Module (SAM) code, and developed Model Predictive Control (MPC) algorithms for control drum system. This report covers three topics. The first part analyzes the dynamics of the variable flow controllers proposed for a passive reactivity control system. The analysis provides reference results on how much the power can be changed with flow rate change. Our preliminary results show that the flow rate change cannot insert the reactivity instantaneously due to the thermal time constants of graphite. Therefore, the variable flow rate controller is a delayed control system. The MPC algorithm for variable flow rate controller is then developed and analyzed. This step is performed to determine how the flow controller should operate optimally–by assuming we can control it–so that we have requirements for the performance of the passive system that is being designed. Results indicate that the variable flow rate controller can achieve load following for cases where the power ramp rate is smaller than 5% %0/min with error smaller than 0.1% %0. These results also agree with our previous expectations. The major challenge for the passive flow rate controller is that a relatively fast flow rate change is required when the power starts or stops changing. Next we develop a simplified TH model based on a previously developed model for Very High Temperature Reactor (VHTR) designs. The purpose of developing the feedback model is to reduce the computational cost of the TH calculation in the high-fidelity transient neutronics simulation when developing and verifying the control system. The simplified TH model is developed to be consistent with the SAM model in our previous report. The simplified TH model is then verified for a unit cell with the power distribution used in the SAM model. It is shown that the simplified TH model can produce results that agree with SAM results very well with a maximum difference at any point being less than 10 K (which occurs in the graphite). Therefore, we expect that the high-fidelity neutron transport calculation coupled with the simplified TH model to produce reasonably accurate results. Finally, we lay out the plan for implementing the MPC algorithm in the PROTEUS code. The plan includes forming the model, developing the nonlinear MPC, and developing the basic solvers to solve these problems. Future work includes implementing the feedback model into the PROTEUS code, and performing coupled high-fidelity simulation. Studying the nonlinear MPC control to see whether it will improve the performance of the control system is another important topic. Eventually, the MPC will be implemented into the PROTEUS code and be used to verify the performance of the semiautonomous control system.