Annular CoreResearch ReactorInstrumentation & Control Systems Pulse Diagnostics System Upgrade SAND #2007-4211 TRTR Conference 17-20 September 2007 Keith Barrett, PrimeCore David Clovis, Operations Staff Robert Zaring, Engineering Support Staff Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company,for the United States Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000.
Abstract Sandia National Laboratories (SNL) and PrimeCore Systems have developed a pulse diagnostic measurement system for the SNL Annular Core Research Reactor (ACRR) based on PXI instrumentation and leveraging ZTEC ZT410PXI modular oscilloscopes. The power and flexibility of the PXI platform and the ZT410PXI 16-bit, 400 MS/sec instrument enabled SNL and PrimeCore to replace and decidedly improve an aging system based on 8-bit bench-top scopes, amplifiers, and long cable runs. The ACRR allows scientists to perform experiments and test components under extreme radiation exposure. The research reactor is used to perform in-pile experiments for radiation effects, reactor development, and safety requirements. These experiments are conducted by “pulsing” the reactor, which allows for extremely high power levels to be achieved for short periods of time. This power then decays exponentially over time. To ultimately capture the power measurements, Self Powered Neutron Detectors (SPND) are placed just outside of the reactor core. They generate currents that are proportional to the neutron flux of the reactor, which can be related to the reactor power. The system employed by SNL and PrimeCore then amplifies and converts the current to voltage using termination resistors, a PXI current amplifier, and an external programmable current amplifier. This method allows them to capture many different ranges and conditions. Finally, the high-resolution ZTEC ZT410PXI modular oscilloscopes acquire the voltage measurements and transfer the data to a control computer for processing.
Pulse Diagnostics System Upgrade Annular Core Research Reactor • 500 pneumatic operations/yr • 380 reactor pulses/yr • Pulses: 33,000 MW peak power, 7-7.5 msec pulse width, 280-310 MJ (Safety Limit: 500 MJ, 60,000 MW) • Steady State: Limited by license to 2.0 MW (increasing to 2.8 MW in November 2007), have run as high as 4.0 MW.
Pulse Diagnostics System Upgrade • Why upgrade old system? • Replace Aging system • Upgrade Instrumentation • Improve Resolution and Accuracy, Reduce Noise • Improve and simplify user interface • Provide simple system calibration
Pulse Diagnostics System Upgrade • System Output Comparison ~ 640,000 data points ~ 12,000 data points
Previous Pulse Diagnostic System Low-RangeSPNDs High-Range SPNDs Termination Resistor Termination Resistor Pulse Timer Trigger Oscilloscope I/V & amplifier Analog/Digital Converter & StorageDiagnostics Computer I/V & amplifier
Key Hardware Improvements • Reduced Cable Run • Instrumentation is located in High Bay. • Cable run is reduced, reducing system noise and capacitance. • Fiber optic connection from instruments to computer in control room. • Improved Digitizers • 8 bit scope replaced by 16 bit digitizer. • Digitizer accuracy approximately 0.25% compared to >2%. • Modular Instruments, Installed spare. • Log Amplifiers • Converts 10 na to 10 ma to .4 to 1.4 Volts • Provides even resolution across 6+ decades of range. Approximately 5000 bits\decade of resolution compared with 512 bits (in 2 separate ranges) of resolution in the old system.
Key Software Improvements • Integrated System • Single application runs the system, processes the data, views and manages data. • System calibration \ verification is built into the software package. • Excel is used as a post analysis and printing tool. It is not used for signal processing and calculation. • Predictive Calculations embedded in program. • Source Code Control and Installation • System will be an executable program with an installation package for recovery. • Algorithms will be “locked” with the executable. Can only be changed with a new revision and release.
Software Design\Implementation • Simple Architecture, Clean Program. Basic event driven interfaces with simple state machines. • Object based data flow. • Modular programming. • Basic Documentation in Code • Source Code Delivered to ACRR • LabVIEW 7.1
Pulse Diagnostics System Upgrade • Future Work • Allow for measurement in Transient Rod Withdraw Submode. • Obtain better results for Steady-State operations <5% Rx Power. • Procure new larger SPND, install, and configure.
End result: Pulse Diagnostics Operational!!! A job well done to all involved!!!
Questions??? How much did it cost? $105K - significant cost savings in not immediately adding new larger volume SPNDs. Plans to upgrade to a large Cad will occur after the fact and cost another $60K. How long did it take? - 4 months to become operational, testing and tweaking is still ongoing as we have time. How much regulator involvement was required? - Not much, since it's not a safety system it required little oversight. No control functions are dependant on pulse diagnostics. Our wide-range and PPS channels still control shutdowns & scrams. What is benefit to experimenters? - Dramatically improved resolution of pulse energy versus time, significantly more flexibility in processing data, and the ability to reprocess raw data from prior tests (of the new system). Data output and detector sharing flexibility? - Any detector can be aligned with any amplifier fairly easily. Multiple choices for output, including standalone data viewer where an experimenter can reprocess the raw data from his desktop. Why is ‘pulse’ diagnostics used for ‘steady-state’ operations? - By using the same detectors we can achieve decent power trace resolution at ~5 % (120 kW) which allows for long-term plotting of SS data and associated parameters. We expect significant improvement (i.e. able to trace power at much lower levels say 1.2 kW) with the new large Cad detector. How difficult is ‘power’ calibration for pulse detectors? - Calibration from the detector input to the PC is a simple process using a current source and the operating program. The program controls the current source running through a prescribed routine. Calibration of the actual pulse vs. the measured pulse is a ‘whole nother story’. We have no idea what our pulses REALLY look like. Input impedance is a large factor in trying to trace power from ≤ 1 kW to ~33 GW by measuring current. We "back calibrate" our pulse height based on dosimetry data of overall MJ based on theoretical pulse shapes. Can the ‘new’ system interpret ‘old’ system data? - YES. It can also compare data over multiple runs. We cannot reprocess old system data since the raw inputs were not saved. Are there unique features of the new system? The ability to integrate reactor energy during any part of the pulse is a built in flexibility of the new system compared to old system. Some experimenters only care about integral energy during portions of the pulse (say immediately after a circuit is energized). Are there any cost savings of the new system? The new system saves calibration costs (~$1200 per year) of old scopes.