ML25155A009

Us Dept of Defense, Uniformed Services Univ of the Health Sciences - Follow Up Report for Event Notification (EN 57730)

ML25155A009
Person / Time
Site:	Armed Forces Radiobiology Research Institute
Issue date:	06/04/2025
From:	Burke G, Smolinski A US Dept of Defense, Armed Forces Radiobiology Research Institute
To:	Office of Nuclear Reactor Regulation
Shared Package
ML25155A007	List: ML25155A008 ML25155A009
References
EN 57730
Download: ML25155A009 (1)

Text

Follow up Report for Event Notification (EN 57730)

Armed Forces Radiobiology Research Institute (AFRRI)

TRIGA Mark-F To satisfy the requirements of:

U.S. Nuclear Regulatory Commission License No. R-84, Docket No. 50-170 Technical Specification 6.5.2 Prepared by:

Mr. Andrew Smolinski Reactor Facility Director, AFRRI Submitted by:

Gerald F. Burke CAPT, MSC, USN Institute Director, AFRRI

AFRRI - EN57730 Follow-up Report

Page 2 of 24

Table of Contents

1. Executive Summary

2. Introduction & AFRRI General Information

3. Initial Event Notification

4. Circumstances Preceding the Event

a. Nuclear Instrumentation Description

b. Automatic Control Algorithm

c. Experiment in progress

d. Reactor Operations in Support of Experiment

5. Event Sequence and Severity

6. Direct Cause of the Event

7. Current Effects on Facility and Corrective Actions

8. Status of Long-Term Corrective Actions

9. Summary

Attachments:

Attachment A: Memorandum for Reactor Staff, Training, Temporary Restrictions, and Path Forward Following High Power Scram, dated 27 May 2025.

Attachment B: Scram 0027 Scram Report Form Attachment C: Corrective Maintenance CM 25-015 NLW Noise Troubleshooting

AFRRI - EN57730 Follow-up Report

Page 3 of 24

1. Executive Summary At 1055 EDT on 22 May 2025, the AFRRI TRIGA Mark F (License: R-84, Docket: 50-170) experienced a high-power SCRAM (emergency shutdown) during startup in steady-state mode under two (2) control rod automatic control. During power escalation, the Logarithmic Nuclear Instrument Channel (NLW) malfunctioned, failing to provide a period signal (rate of change of power) to the automatic control system. The apparent cause of the event was high frequency noise induced in the log channel detector input, causing the channel to fail to a fixed high frequency of constant amplitude. The flat signal produced a negligible period signal for input to the automatic control algorithm. With no significant period input, the automatic control algorithm produced an output that withdrew the two control rods continuously for 5-7 seconds.

The reactor power level increased rapidly, several orders of magnitude over ~5-7 seconds, until high flux safety channels 1 and 2 caused an automatic SCRAM, releasing control rods to safely shut down the reactor. Due to the rate of power increase, reactor power momentarily exceeded the steady-state license limit of 1.1MW for approximately 100 msec. Peak power level recorded by the digital chart recorders was 1.2MW. All reactor protection systems functioned as designed. Reactor power was verified as decreasing; control rods were verified to be fully inserted. Neither the fuel temperature Safety Limit (1000 °C), nor the Limiting Safety System Setting (600 °C), were exceeded during the event. Maximum fuel temperature recorded was 36

°C. All radiation monitoring systems were operable during the event. No release of radioactive material was detected. No damage to the fuel occurred.

The event was reported to NRC Headquarters Operations Office at 1504 on 22 May 2025 (EN 57730).

Subsequent troubleshooting revealed that the source of the noise is related to AC-grounds. The largest sources that cause a change in signal are:

Switching the pool lighting, Manipulating the transient rod drive AC/DC motor circuits, Movement of the core, particularly at the region transition points, Manipulating the NLW detector flexible conduit.

Operators were briefed on the event with increased emphasis to be aware of these activities, indications, and failure modes during operation. Based on results of troubleshooting, minor flexing of the NLW flex conduit away from the transient rod drive wiring appears to restore the indications to normal.

The following short-term restrictions were put in place:

Minimize use of 2-rod automatic for routine automatic startups until auto mode drive speeds can be tuned properly (FSM-003 & FSM-005). Since the period interlock is bypassed during square-wave ramp-up, use of 2-rod auto for square wave operations is still authorized.

AFRRI - EN57730 Follow-up Report

Page 4 of 24 If indications of NLW noise become apparent during an automatic startup, immediately place the system in manual, and dispatch an operator to the bay to immediately correct the issue in accordance with Tech Spec 3.2.1.a.(2).

The following longer-term actions are being pursued:

Transient rod drive motor modification to use a stepper motor similar to other drives rather than AC circuits switching a DC motor (FM-031 & FSM-005). This modification was awarded to General Atomics on 20 Sep 2024 with an anticipation completion date no later than 31 Dec 2025.

Drive speed tuning for various modes of operation and number of drives being controlled in automatic to reduce the reactivity insertion rate (FSM-003 & FSM-005).

This modification in part has been awarded to Plantation Productions (OEM software vendor) on 12 Mar 2025 and will be implemented in conjunction with the transient rod drive motor modification with an anticipation completion date no later than 31 Dec 2025.

Investigation of the condition of building AC ground network and potential presence of shared neutrals, and modification of electrical distribution system to improve grounding based on results of the investigation.

Fission chamber, cabling, jacket, and connector replacement on NLW once materials can be procured. The replacement is scheduled during the upcoming annual maintenance outage in November/December 2025, but no later than 31 Dec 2025.

2. Introduction and AFRRI General Information

The Armed Forces Radiobiology Research Institute (AFRRI) is located on the grounds of the Naval Support Activity Bethesda (NSAB), Bethesda, Maryland. AFRRI is a tri-service military organization under the Uniformed Services University of the Health Sciences. AFRRI conducts research in the field of radiobiology and related matters that are essential to the operational and medical support of the Department of Defense. AFRRI is staffed by members of the three military services and by civilian personnel. In support of its mission, AFRRI operates a research reactor for the generation of a large mixed-field (neutron and gamma) radiation source.

The non-power research reactor is a General Atomics TRIGA Mark F reactor; TRIGA is an acronym for training, research, isotopes, General Atomics. The AFRRI TRIGA Mark F can operate in steady state or pulse mode. When the reactor is in the steady-state mode, control rods are withdrawn from the reactor core until the desired power level is achieved. In the pulse mode, one of the control rods is rapidly ejected from the reactor core by using a special compressed-air control rod drive mechanism. The reactor is licensed for a steady-state power level of 1.1-megawatts, but can pulse to 4000 MW, with a step reactivity insertion limited to $3.50 in pulse mode by the technical specifications. With the rapid ejection of a relatively large amount of neutron absorber from the core, the fission chain reaction escalates at a rapid rate, called a

AFRRI - EN57730 Follow-up Report

Page 5 of 24 prompt critical excursion. As the fuel heats during the pulse, the inherent effects of the uranium zirconium hydride (UZrH) reactor fuel rapidly bring the reactor down to a low power level without human or equipment intervention. The reactor core is suspended under 16 feet of water from a carriage just above the reactor pool. The pool is an effective radiation shield, so personnel can safely observe the reactor as it operates. The unique feature of the TRIGA Mark F is that the carriage rides on a track that allows movement of the core from one exposure room to the other exposure room, as shown in Figure 1. Movement of the core carriage along the track from one side of the pool to the other is controlled from the reactor console. In this configuration, all nuclear instrument detectors, control rod drive mechanisms, core positioning motor, instrumentation, and cabling must travel with the core. Approximately 5 minutes are required to move the core from one side of the pool to the other, a distance of 13 feet. An additional 10 minutes is required to close the lead shield doors in the center of the pool and open the applicable exposure room plug door to allow personnel access to retrieve irradiated specimen. The advantages of a movable core are (a) the quantity and character of the radiation reaching the exposure facilities can be controlled, and (b) more than one exposure facility can be used during a day of reactor operations. Typical experiment operations involve batch irradiations of specimen positioned in the exposure room with multiple core movements and startups in a given work day.

The TRIGA reactor is designed to be inherently safe, i.e., the reactors design is such that there is no possibility of an accident that can produce an unsafe condition for the staff or the general public. Since the late 1950s, more than 70 TRIGA reactors have been constructed worldwide and used for a variety of applications with no accidents. The TRIGA reactor at AFRRI has been operating since 1962 and is the only remaining operational Mark F design in the world.

Figure 1. Cutaway view of AFRRI TRIGA reactor

AFRRI - EN57730 Follow-up Report

Page 6 of 24

3. Initial Event Notification At 1055 EDT on 22 May 2025, the AFRRI TRIGA Mark F (License: R-84, Docket: 50-170) experienced a high-power SCRAM (emergency shutdown) during startup in steady-state mode under two (2) control rod automatic control. During power escalation, the Logarithmic Nuclear Instrument Channel malfunctioned, failing to provide a period signal (rate of change of power) to the automatic control system. The reactor power level increased rapidly, several orders of magnitude over ~5-7 seconds, until high flux safety channels 1 and 2 caused an automatic SCRAM, releasing control rods to safely shut down the reactor. Due to the rate of power increase, reactor power momentarily exceeded the steady-state license limit of 1.1 MW. Peak power level recorded by the digital chart recorders was 1.2 MW. All reactor protection systems functioned as designed. Reactor power was verified as decreasing; control rods were verified to be fully inserted. The fuel temperature Safety Limit (1000 °C), nor the Limiting Safety System Setting (600 °C), were exceeded during the event. Maximum fuel temperature recorded was 36 C. All radiation monitoring systems were operable during the event. No release of radioactive material was detected. No damage to the fuel occurred.

This event was reportable to the NRC under AFRRI Technical Specifications 6.5.2.b. (Operation in violation of any Limiting Condition for Operation (LCO)) and 6.5.2.c. (Malfunction of a required safety system component during operation that renders the system incapable of performing its intended safety function).

LCOs affected:

T.S. 3.1.1. - Steady state power level exceeding 1.1 MW T.S. 3.2.1.a. - Log Channel inoperable T.S. 3.2.2. Table 3 - Failure of Rod Withdrawal Interlock for a period less than three (3) seconds.

The event was reported to NRC Headquarters Operations Office at 1504 on 22 May 2025 (EN 57730).

4. Circumstances Preceding the Event Reactor Control System: The entirety of the reactor control system, nuclear instrumentation, and facility interlock system was replaced from 2016-2018, and approved by the NRC for operation on 22 June 2022 with reactor license R-84 Amendment 26. The Reactor Instrumentation and Control (I&C) system is a hybrid computer-based system which includes a hardwired Reactor Protection System (RPS) with dedicated displays and controls so that safe operation and monitoring of the reactor can continue should the computers become unavailable.

The primary function of the RPS is to scram the reactor by allowing the control rods to fall into the core in response to automatic protective actions or actions initiated by the operator from the Control System Console (CSC) operator interface in response to other abnormal reactor operating conditions that may arise during the course of operations. The equipment installed in the Data Acquisition Cabinet (DAC) in the reactor bay near the pool acquires data in the form of electronic signals from instrumentation in the reactor and auxiliary systems, processes it, and

AFRRI - EN57730 Follow-up Report

Page 7 of 24 transmits it to the operator via multiple displays on the CSC.

Nuclear Instrumentation Design: The Reactor I&C system receives input from various detectors and sensors. These include at least one fission chamber, a compensated ionization chamber, two uncompensated ionization chambers, and thermocouples embedded in the instrumented fuel elements to measure fuel temperature directly. Signals from these units are processed in the DAC which is housed in the reactor bay. The neutron flux levels are measured from subcritical source multiplication range through licensed maximum power range.Four independent power measuring channels are provided for a continuous indication of power from subcritical neutron source multiplication range to the maximum steady-state licensed power level. Since not all of the neutron flux instruments are capable of this, continuous indication is ensured by maintaining a minimum of one decade of overlap in indication while observation is transferred from one instrument channel to another. The four channels are the wide-range logarithmic nuclear instrument channel (NLW-1000), the Multi-range linear nuclear instrument channel (NMP-1000), and two steady-state high flux channels (NP-1000 and NPP-1000). The NPP-1000 also includes the ability to detect and capture the power excursion during pulse mode operation. The NLW and NMP provide signals to the automatic control algorithm. The NP and NPP provide high flux trips to the reactor protection system, and successfully scrammed the reactor during the event. The NLW, the primary cause of the malfunction, is detailed below.

The NLW-1000 is designated as the Log Power Channel. The design function of the NLW-1000 is to measure neutron flux in order to provide the following:

Wide range logarithmic power indication, i.e. from source range to full power.

Reactor period indication.

Bistable trip/signal for interlocks.

Analog outputs to the bargraphs and recorders for steady-state operation.

Digital outputs to the reactor control console for steady-state operation.

The NLW-1000 monitoring channel is a wide range logarithmic nuclear instrument that operates with a fission chamber and a PA-1000 preamplifier that decouples and amplifies pulses that originate at the fission chamber. The module combines count rate and current measuring techniques to provide continuous power measurement from the subcritical source multiplication range through full licensed power of 1.1 MW. In the lower 7 decades, the NLW-1000 counts pulses coming from the PA-1000, converting the frequency to a voltage output to a summing amplifier, calibrated for 1 volt per decade for up to 7 volts. For the upper 3 decades where individual pulses cannot be resolved, the current draw on the power supply flowing in the fission chamber is monitored and converted to voltage, calibrated for 1 V/decade for up to 3 volts. The summing amplifier combines the two inputs for an overall 0-10 V signal. The transition from pulse to current mode happens automatically. The combination of the pulse count and the current signal provides a continuous logarithmic indication of the reactor power. To prevent higher frequencies from saturating the PA-1000 and affecting the summing amplifier output, the voltage output signal is clamped (e.g. fixed) at approximately 7 V for the frequency section and corresponding indication of approximately 0.1% power (1 kW). The precise setting of the

AFRRI - EN57730 Follow-up Report

Page 8 of 24 clamping circuit (R3) is set during the bench calibration to align both the count rate and current sections for a smooth transition.

It is at this ~7 V value that the instrument failed to during the high frequency noise event of 22 May 2025, an indication of approximately 0.1% power (1 kW).

Figure 2. NLW-1000 Functional Diagram showing relationship between reactor power and summing amplifier output. (Ref. GA Manual T3322000-1UM)

The logarithmic reactor power signal is monitored by a period circuit which generates an output proportional to the rate of change in reactor power at any given instant. This signal called period is a measure of the time (in seconds) it takes for the reactor power to change by a factor of the mathematical constant e (2.718). The period indication is from -30 seconds to +3 seconds. A value of infinity () represents a constant unchanging power level.

The NLW-1000 relies on analog signal processing (no software) of the detector signal for both the power signal and the period signal, along with the bistable trip activation. The NLW-1000 also provides analog outputs to the bargraphs and chart recorder for use at the reactor control console. An onboard analog-to-digital converter provides data to the CSC over ethernet.

Historically, noise has been a problem since initial installation of the log channel in 2017.

Several incremental improvements have been made since reactor restart in 2022.

AFRRI - EN57730 Follow-up Report

Page 9 of 24 In July 2022, initial startup revealed substantial noise being generated in the cable from the detector to PA-1000 due to proximity to the control rod drive cables. The transient rod drive cable carries 120 VAC voltage to switching relays to switch the direction of a DC motor mounted to the drive stand. The other three drives carry frequency signals between the stepper motors and driver modules in the DAC cabinet. This was addressed by rerouting those cables through the center of the metal cable boom, providing some separation and electrical shielding and substantial improvement in signal quality. During the same maintenance activity substantial quantities of abandoned cable was removed from the cable troughs and detector cabling was rerouted away from AC circuits as much as possible.

120 VAC noise from the dolly motor was addressed in June 2023 during a maintenance outage by separating and securing the AC circuit wires in a flexible metal conduit, secured to the cable boom away from detector cables, which were hung below the boom. At this time the cable jacket material on the detector cable was also insulated using heat shrink sleeving to minimize the potential for contact with the metal detector housing as the core moves.

Further separation was achieved by separating the PA-1000 preamplifier into its own cabinet.

In 2024, the cable from the detector chamber housing on the core shroud to the PA-1000 was routed in a dedicated flexible metallic conduit along the boom, providing additional electrical shielding and separation.

In 2024, a hardware modification (FM-031) was proposed to convert the transient rod drive system to a stepper motor configuration to eliminate this source of AC switching noise. This was approved for FY-25, and contract HU000124F0067 Task Order 3 was awarded on 20 September 2024 to General Atomics to begin design work. The change also requires a facility software modification to the reactor control system software (FSM-005). The software modification portion of the project was awarded to the original OEM software vendor, Plantation Productions on 12 March 2025. The anticipation completion date for the project is no later than 31 December 2025.

Automatic Control Algorithm: The reactor control system, when placed in Automatic Mode, will automatically control positions of the Shim, Safety and Reg rods, depending on bank selection, to maintain a specific power level based on the % Power reading from the NMP-1000, the reactor period from the NLW-1000, and the demand power level. The demand power level is taken from the setting of the power demand setpoint on the Left Side Status display and indicated at top left of the Right-Side graphics display on the user interface terminal (UIT) computer displays. The inputs are collected in the Central Control System (CCS) computer, where the software algorithm calculations are made and rod speed and direction commands are generated.

In Automatic Mode, the computer controls the rods based on the bank selection according to a proportional derivative (PD) algorithm to drive the rods either up or down based on a comparison of the reactor power with the demand power and reactor period. Any trip/warning from the NMP-1000, NLW-1000, or a communication error from either instrument will result in rod control reverting back to Manual Mode. Automatic Mode must be re-enabled once condition(s) clear.

AFRRI - EN57730 Follow-up Report

Page 10 of 24 In automatic mode, the following rod combinations can be selected for banked rod movement:

Regulating Rod Only Shim and Regulating Rod Shim, Safety, and Regulating Rod When utilizing the automatic mode, the reactor power (from the NMP) is compared against the power demand setting to obtain an error number. This power mismatch error is combined with a second error term generated based on the period signal (from the NLW). When the power demand has more than 2% deviation from the measurement from the NMP-1000, the rods (as selected by the combinations listed above) are moved into or out of the core. Rods are controlled with variable speed to allow for minor corrections in reactor power with small deviation and major corrections in reactor power for large deviations. Variable speed also enables the algorithm to achieve the desired power with minimum overshoot or undershoot. The rod speed may never exceed hardwired limit of 30 inches/minute. The NLW-1000 period signal is provided and functions to limit the rate of power increase to a period of +6.5 seconds by decreasing rod speed as period approaches +6.5 seconds or inserting rods when at a period less than +6.5 seconds. The NLW-1000 will inhibit rod withdrawal if the reactor period exceeds +3 seconds in manual and automatic modes. Control rod position is denoted in units on the rector control console, where 0-1000 units is linearly scaled for the full span of withdrawal (15 inches).

The automatic control algorithm is described for operators in Procedure 002 as follows:

The control algorithm functions based on the Five Rules of Automatic Mode:

1. If the reactor is within +/-2% of demand power, Automatic Mode does not move rods. This is called the dead band.

2. A 3 second period will always mandate max rod speed insertion unless in the dead band.

3. Any mismatch >20% will act like a 20% mismatch, providing a cap on the magnitude of the mismatch term.

4. For a mismatch less than 20%, the rod speed movement is scaled linearly so the rod will travel slower as the power gets closer to demand power.

5. A term based on period is used to buck, i.e., buffer, the mismatch term and is scaled inversely with the period signal, i.e., the faster (shorter) the period the slower the rod speed. On a large up-power (>20%), the period term will overwhelm the mismatch term at ~10 second period. Within 20% mismatch that control point is scaled back as you get closer to demand.

In the event of 22 May 2025, the fifth rule ultimately was the cause of the rapid rod withdrawal.

With no significant period signal, due to the flat NLW power signal, there was no buffering the power mismatch error term, which was at its maximum contribution to the algorithm.

Square Wave mode is a special operating mode which allows large, rapid reactivity insertions less than $1.00 followed by a switch to automatic control to achieve steady state power.

Operating in Square Wave mode must begin with the reactor in manual mode. With the power less than 1 kW (as determined by the NLW-1000) and the transient rod air supply turned off, the

AFRRI - EN57730 Follow-up Report

Page 11 of 24 Square Wave mode switch can be depressed. This will change the console from manual to Square Wave mode. Upon pressing the FIRE button, the reactor power will increase to the demand power level and inhibit control rod outward movement during power escalation. Upon achieving the desired power level, the console will switch to Automatic Mode to maintain the reactor at this constant power level using the automatic control algorithm as described above. In square wave operation, automatic control effectively begins at or near the target power, with small control rod adjustments to counteract the rising fuel temperature. When performing square waves from critical, a minimum of two control rods may be required to provide enough positive reactivity to negate these negative reactivity effects. If the desired power level is not reached within 30 seconds, the system will switch to Manual Mode and display a message to the operator on the Annunciator Pane. Since automatic control is not triggered in square wave mode until target power level is reached, the NLW signal is well past the frequency portion of the instrument, so the failure mode that led to this event is not a factor in square wave operation.

Experiment in Progress and Reactor Operations in Support of Experiment Campaign: The experiment in progress on 22 May 2025 involved several batch irradiations of specimen in Exposure Room 1 (ER1). The Reactor Use Request (RUR 25-028) was a routine experiment campaign which called for 5 separate batch exposures at 15% neutron contribution for the total mixed field neutron-gamma (n-g) dose, at 5 separate total n-g dose targets, at a 0.6 Gy/min total n-g dose rate. The specimen are limited to 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> in the exposure room. The campaign would require a total of 6 reactor startups and 12 core movements. To reach the 15% neutron percentage, ER1 is configured with no supplemental shielding (bare configuration), the specimen table was positioned to 205 cm, and the target core location was position 264 (equating to 8.9 cm back from the region 1 full in position). To minimize and match as close as possible the same ramp time and dose to specimen for each batch, automatic control is used to ramp the reactor to the target power level, and a scram and simultaneous core move away from ER1 is used to end the exposure. The timeline for the experiment campaign was planned and executed as follows:

0600-0630 Open ER1 plug door. Install ion chambers for dosimetry setup run and prepare dosimetry data capture equipment. Close ER1.

0630-0711 Perform daily startup checklist and associated surveillances.

0711-0725 Open Lead Shield Doors. Move core to position 500 (Region 2). Perform reactor startup and k-excess surveillance.

0725-0752 Move core to position 264 (Region 1). Dispatch second operator to verify -8.9 cm core position in bay. Increase reactor power to target power level (~15 kW). Capture dosimetry data and pinpoint target power level and target rod positions for desired dose rate for exposures.

0752-0800 Perform reactor shutdown. Move core to position 700 (Region 3). Close lead shield doors.

0800-0845 Open ER1 plug door. Remove ion chambers. Install specimen for first exposure.

Close ER1.

0845-0906 Open Lead Shield Doors. Move core to position 264 (Region 1) while manually positioning 2 control rods at target position and 2 control rods to the predetermined subcritical position. Dispatch second operator to verify -8.9cm core position in bay.

AFRRI - EN57730 Follow-up Report

Page 12 of 24 Increase reactor power to target power level in automatic (2-rod automatic - 15.2 kw).

Hold power until exposure at target dose.

0906-0915 Manually scram reactor. Move core to position 700 (Region 3). Close lead shield doors.

0915-0929 Open ER1 plug door. Install specimen for second exposure. Close ER1.

0929-0950 Open Lead Shield Doors. Move core to position 264 (Region 1) while manually positioning 2 control rods at target position and 2 control rods to the predetermined subcritical position. Dispatch second operator to verify -8.9 cm core position in bay.

Increase reactor power to target power level in automatic (2-rod automatic - 15.2 kw).

Hold power until exposure at target dose.

0950-0954 Manually scram reactor. Move core to position 700 (Region 3). Close lead shield doors.

0954-1006 Open ER1 plug door. Install specimen for third exposure. Close ER1.

1006-1028 Open Lead Shield Doors. Move core to position 264 (Region 1) while manually positioning 2 control rods at target position and 2 control rods to the predetermined subcritical position. Dispatch second operator to verify -8.9 cm core position in bay.

Increase reactor power to target power level in automatic (2-rod automatic - 15.2 kw).

Hold power until exposure at target dose.

1028-1032 Manually scram reactor. Move core to position 700 (Region 3). Close lead shield doors.

1032-1046 Open ER1 plug door. Install specimen for fourth exposure. Close ER1.

1046-1055 Open Lead Shield Doors. Move core to position 264 (Region 1) while manually positioning 2 control rods at target position and 2 control rods to the predetermined subcritical position. Dispatch second operator to verify -8.9 cm core position in bay.

Increase reactor power to target power level in automatic (2-rod automatic - 15.2 kw).

1055 Reactor scram - High flux trips on NP and NPP. Verified scram indications and made notifications to reactor operations supervisor (ROS) and reactor facility director (RFD).

1055-1150 Move core to position 700 (Region 3). Close lead shield doors. Open ER1 plug door. Retrieve specimen. Close ER1 1205 ROS decision made to postpone partial exposure run and last exposure pending resolution. RFD concurred.

1507 Initial Event Notification (EM 57730) made to NRC.

5. Event Sequence and Severity The reactor control system includes a feature called history playback, which captures all events written to the operator displays and records them to a file on the computer for future playback.

This data history-logging automatically begins whenever the operator resets the scrams (a prerequisite to reactor startup). Data is recorded approximately every 100 ms. Also, any time there is a change in the WARNINGS pane or the SCRAM pane, that event is recorded, regardless of time. Logging continues following a system scram until terminated by the operator by turning the keyswitch off.

AFRRI - EN57730 Follow-up Report

Page 13 of 24 History playback takes the form of replaying snapshots of the two video display screens (Right Side Reactor Displays and Left Side Status Display) in sequence as they were recorded.

Whenever a value or graphic changes on either of the display screens, that new value is written to the history playback file; writing only the display differences helps reduce the size of the playback file. During playback, the operator will see the control rod movements, trends, and bar graph displays as they occurred during real time operation. When played back the data can be displayed automatically in sequence at a variable speed, allowing the viewer to slow down the speed for detailed analysis. The history playback data displays along with the raw data text files were analyzed for event reconstruction.

As described in the preceding section, 4 startups and 3 successful exposures in the exact same sequence were performed prior to the event. During the fifth startup (Exposure 4) of RUR 25-028, while bringing the power level to 15.2 kW the reactor period indicator did not respond to automatic rod withdrawal. The reactor was at core position 264 (-8.9 cm) with the four control rods prepositioned at 592 (transient), 254 (shim), 591 (safety), 252 (regulating) in preparation for 2-rod auto up-power to 15.2kW. At 10:55:01, auto mode was initiated. At 10:55:25, the Reactor SCRAMMED due to NP/NPP high power trip. Power indications verified that the reactor protection system SCRAMMED as required. NFT fuel temperature trips did not occur. No radiation alarms or other abnormal alarms or conditions occurred.

A detailed review of the history playback files on the console was conducted for the runs on 22 May 2025. The first 3 startups showed no abnormal indications. The fourth startup history file indicates that the NLW channel started behaving abnormally at 10:33 after the reactor was shut down from the previously completed experiment (RUR 25-028, Run 4, Exp 3) after the core was moved from Region 1 to Region 3 (~4 minutes after shutdown) and the lead doors began shutting. Analyzing Figure 3 below shows that the channel noise can be affected by core movement. The NLW indication spontaneously increased to 1-2x10-2% (Figure 3). However, at this point in the shutdown sequence the operator has removed the key, secured the reactor, and is on his way to the prep area to support exposure room operations. The signal remained slightly elevated until approximately 10:51 when the next startup began. Since the reactor key has been removed, the data between these times is not recorded in history playback.

AFRRI - EN57730 Follow-up Report

Page 14 of 24

Figure 3 - Rx shutdown after Run 4, Exp 3 In the fifth start up history file, prior to the event, the NLW showed some noise as the transient rod was prepositioned in manual for startup, indicating values above source range (Point 1 of Figure 4). The transient rod and safety control rods were pre-positioned at the expected 15.2 kW target positions (~590 units). The shim and regulating control rods, which would be controlled in automatic for the power escalation were pre-positioned in manual to ~250 units. This is a known subcritical condition since criticality would be expected above over 500 units based on the cold clean excess reactivity measurement performed during the first startup. During the pre-positioning, the core was moving across the pool to position 264, and nuclear instrumentation readings are expected to increase due to subcritical multiplication from the control rod withdrawal, which they did (Points 12 of Figure 4). The subcritical control rod pre-positioning during core movement is intentional, to minimize the ramp time to target power once in final position, which minimizes the uncertainties in the dose to the experiment in the exposure room. At target power, the control rods would end in a banked configuration maintaining the target power with no operator intervention. This repeatable sequence also minimizes variations between experiment batches.

At 10:55:00, the reactor was placed into automatic mode with NMP power indicating 5.1 mW and a target power of 15.2kW (Point 2 of Figure 4). The position of control rods are Transient (592), Shim (254), Safety (591), and Regulating (252). Based on the k-excess measurement earlier in the day, the reactor is subcritical by $2.41. The system begins to withdraw control rods. At the same time, final core position is being set and visually verified locally by a second operator in the bay. At approximately 10:55:06, due to high frequency noise the log channel flat-lined at 2.07E-2% (Point 3 of Figure 4). At approximately 10:55:17 the Shim and Regulating control rods reach 569 units and 560 units respectively and linear power (NMP) indication

AFRRI - EN57730 Follow-up Report

Page 15 of 24 begins to rise appreciably (Point 2 of Figure 4). Based on the k-excess measurement, a Shim rod and Reg rod position of approximately 566 units, would be an expected critical condition, however, with the NLW providing a false constant and unchanging indication it is failing to provide significant period feedback. At this point the interlock that would prevent rod withdrawal for a period in excess of 3 sec is not functional, and the target of 6.5 sec (automatic control rod algorithm control point) is not effective. The region between Points 3 and 4 of Figure 4 show the response of the automatic control algorithm; the control system continues to pull the Shim and Regulating rods at full speed as there is no feedback from the log channel to buck or oppose rod movement, with multiple rapid range changes on the NMP over the next ~7 seconds. The shim and regulating control rod maximum withdraw was 720 and 710 respectively before the algorithm sees a fast period and responds.

Figure 4 - Left Side Reactor Display for Run 5, Exp 4 Two UIT software update cycles (~200 ms) prior to activation of the reactor safety system (10:55:25), the automatic control system reacted to a high period (6.9 sec) and power level past the target (provided by the NMP) and began to reinsert control rods (Point 4 of Figure 4). At this point, the NLW had crossed into the current region of the instrument, where frequency input is no longer a factor, and began reading correctly. This control rod insertion attempt is evidenced by a change in SHIM rod position changing from 720 to 717 and both SHIM and REG rod DOWN indicators highlighted on the console and commands appearing in the text file.

Unfortunately, the response was not fast enough. The reactor protection system responded as designed.

1 3

2 5

4 6

AFRRI - EN57730 Follow-up Report

Page 16 of 24 Both high flux safety channels, NP and NPP, tripped at 10:55:25 (console time) which triggered the reactor SCRAM and dropped rods (Point 5 on Figure 4). The region between point 5 and 6 shows the reactor power decay after 5 minutes and a restored log channel signal. No manipulation of the hardware was performed. The core was moved back to Region 3 and the Pb door was closed to allow for entry in Exposure Room 1 to retrieve the experiment.

Evaluation of the console historical log shows the NP and NPP both triggered the SCRAM (Trip

1) immediately upon sensing a signal >108 A (1.08 MW) before the analog data could be digitized and processed for display, indicating the over-power condition lasted less than one (1) update cycle of the UIT console (<100ms). Trip 2 (High Power Alarm 1.05 MW), as well as Trips 3 and 4 (Status Indication Only 1.20 MW) also registered on both the NP and NPP. Since these trips latch, the indications were recorded in the history playback file, providing the evidence that power exceeded the license 1.1 MW steady-state limit. Maximum power recorded by the Honeywell chart recorders was 119.85% (~1.2 MW) as indicated by the multi-range linear channel (NMP). All reactor protection systems functioned as designed. Reactor power was verified as decreasing; control rods were verified to be fully inserted. The fuel temperature Safety Limit (1000 °C), nor the Limiting Safety System Setting (600 °C), were exceeded during the event. Maximum fuel temperature was recorded by NFT-1 as 36 °C, a 15 °C rise from before the event. The NLW Channel returned to normal operation after the event (Point 6 on Figure 4). All radiation monitoring systems were operable during the event. No release of radioactive material was detected. No damage to the fuel occurred. Based on the in-hour equation and analysis of the power and time stamps in the history file, the reactivity inserted above criticality was <$1.00. The maximum step insertion allowed by technical specifications is

$3.50, and the maximum step insertion analyzed in the SAR is $4.00. A scenario involving failure of the 3-second interlock during a more limiting 3-rod automatic control power escalation from low power was specifically analyzed in License Amendment Request (LAR) 26 for approval of the new control system. The resulting $1.19 insertion was terminated by the reactor protection system, and results were still bounded by the $4.00 step insertion SAR analysis.

This event was reportable to the NRC under AFRRI Technical Specifications 6.5.2.b. (Operation in violation of any Limiting Condition for Operation (LCO)) and 6.5.2.c. (Malfunction of a required safety system component during operation that renders the system incapable of performing its intended safety function).

A more detailed analysis of the event indications from the chart recorders and history playback feature can be found in Attachment B.

6. Direct Cause of the Event The direct cause of the event was high frequency noise which affected the NLW-1000 signal while below 1 kW. The high frequency noise caused the instrument to fail to the high end of the lower 7-decades (~1 kW) where frequency determines the output signal to the summing amplifier. As described in Section 3, above 1 kW (0.1%) the current draw on the power supply flowing in the fission chamber is monitored to produce the NLW signal since individual pulses

AFRRI - EN57730 Follow-up Report

Page 17 of 24 from the preamp cannot be resolved. Output from the frequency section is clamped and held constant, so high frequency noise no longer has an impact.

Corrective maintenance item (CM 25-015) was issued on 23 May 2025 to troubleshoot and identify source(s) of the intermittent noise. CM 25-015 is attached as Attachment C to this report. Troubleshooting revealed that the source of the noise is related to AC-grounds. The largest sources that cause a change in signal are:

Switching the pool lighting, Manipulating the transient rod drive AC/DC motor circuits, Movement of the core, particularly at the region transition points, Manipulating the NLW detector flexible conduit.

Testing strongly indicates an intermittent grounding issue within the NLW flex conduit located between the drive support upright and the cable trunk. The NLW signal cable is housed inside a shielded jacket which is housed inside the flex conduit. It is postulated that either the jacket or the signal cable may have a small break in the jacketing leading to intermittent introduction of noise into the channel. Each time, mostly minor movements of the NLW flex conduit at the trunkcore connection would reduce the noise and return the signal to nominal levels.

7. Current Effects on Facility and Corrective Actions There is not sufficient supply of the engineered nuclear detector cable RSS-6-105/LE, jacket material, and connectors at the facility to initiate a repair at this time. The original manufacturer is Rockbestos Company, which became RSCC, which was purchased by Marmon Industrial Energy & Infrastructure. The manufacturer along with approved vendors have been contacted to procure additional cable to facilitate the complete replacement of the cable from the fission chamber all the way to the NLW. This engineered cable is manufactured only on an as-needed basis, requires an estimated 8-week lead time, with minimum purchase of 1000 ft.

In the mean-time, a briefing consisting of all licensed operators was conducted to provide a detailed review and training on the event of 27 May 2025. A memo was written to document a summary of the event, results of CM 25-015, and summarize short and long-term corrective actions. This memo (Attachment A), the Scram 0027 Report Form (Attachment B), and CM 25-015 (Attachment C) were reviewed and discussed. Emphasis was made that operators should be aware of these activities, indications, and potential failure modes during automatic operation and be prepared to take manual control as necessary. Based on results of CM 25-015, minor flexing of the NLW conduit away from the transient rod drive wiring appears to restore the indications to normal.

The following short-term restrictions are in place:

Minimize use of 2-rod automatic for routine automatic startups until auto mode drive speeds can be tuned properly (FSM-003 & FSM-005). Since the period interlock is bypassed during square-wave ramp-up, use of 2-rod auto for square wave operations is

AFRRI - EN57730 Follow-up Report

Page 18 of 24 still authorized.

If indications of NLW noise become apparent during an automatic startup, immediately place the system in manual, and dispatch an operator to the bay to immediately correct the issue in accordance with Tech Spec 3.2.1.a.(2).

The following longer-term actions are being pursued:

Transient rod drive motor modification to use a stepper motor similar to other drives rather than AC circuits switching a DC motor (FM-031 & FSM-005).

Drive speed tuning for various modes of operation and number of drives being controlled in automatic (FSM-003 & FSM-005).

Investigation of the condition of building AC ground network and potential presence of shared neutrals, and modification of electrical distribution system to improve grounding based on results of the investigation.

Fission chamber, cabling, jacket, and connector replacement on NLW once materials can be procured. The replacement is scheduled during the upcoming annual maintenance outage in November/December 2025.

8. Status of Longer-Term Corrective Actions Transient rod drive motor modification (FM-031, FSM-005):

As described in section 3, the transient rod drive is being modified to use a stepper motor similar to other drives rather than AC circuits switching a DC motor (FM-031 & FSM-005). The hardware modification (FM-031) was proposed to convert the transient rod drive system to a stepper motor configuration to eliminate this source of AC switching noise. This was approved for FY-25, and contract HU000124F0067 Task Order 3 was awarded to General Atomics on September 20, 2024 to begin design work in October 2024. Revised drawings for wiring modifications for the applicable DAC drawers have been completed and part selection for the drive has been completed. A spare drive from AFRRI was delivered to General Atomics in February 2025 to begin modifications. Factory acceptance testing is anticipated to begin in June 2025.

The change also requires a corresponding facility software modification to the reactor control system software (FSM-005). The changes for the software were contracted to Plantation Productions, the original software engineering subcontractor to General Atomics that programmed the console software for original installation. This contract was awarded on March 12, 2025, with anticipated completion of a beta version to begin testing in August 2025. The drive installation, drawer modifications, and software testing for site acceptance testing are all planned and scheduled to be done in the 2025 annual maintenance outage scheduled for 24 Nov 2025 - 04 Jan 2026.

AFRRI - EN57730 Follow-up Report

Page 19 of 24 Control rod drive speed tuning software modifications (FSM-003, FSM-005): Improvement in the drive speed tuning was identified after a special automatic mode test procedure was executed following Site Acceptance Testing Part 2 (SAT-2) (critical operations) in August 2022.

The SAT-2 did not characterize 2-rod or 3-rod automatic. A new Procedure 293 Automatic Mode Characterization was developed and executed in December 2022 to document observations of performance of the algorithm with reactor feedback. Following this testing 3-rod automatic control was restricted because of excessive overshoot/undershoot and hunting, particularly at low powers. As a result of this testing a software modification was initiated to revise the algorithm constants for better performance. A spreadsheet tool was developed to analyze the anticipated voltage output and polarity of the algorithm, which ultimately controls rod speed and direction for various constant settings. The software modification (FSM-003) documentation and execution was postponed pending the stepper motor conversion project for the transient rod, since this will require an update to the analysis. The status of FSM-003 is tracked as long-term corrective action CA020 in the corrective action program. FSM-003 will be combined and executed with FSM-005.

AC Electrical Distribution System and Grounding Investigation: CM 025-017 documents results of some 120 VAC electrical system panel inspections performed over the last year in the reactor space. Investigations began last year when some outlets for non-critical equipment were found mislabeled in the reactor bay. Circuits with shared neutrals were found, as well as shared neutrals for circuits originating from different electrical panels. These results have been shared with USUHS facilities maintenance, who is coordinating with Naval Facilities Engineering (NAVFAC) to identify the scope of the repairs. Changes to the 120 V electrical distribution system must be controlled through NAVFAC for all base facilities.

A separate project to provide an isolated clean 120 VAC power system from the emergency power panel to new vital equipment (radiation monitoring) is being explored. The status of installation of a large uninterruptible power supply (UPS) with AC/DC/AC converting and rectifying capability for filtering was submitted on April 10, 2025 and is currently tracked by facilities with request number RITM0069860 and work order 2505889. The reactor is tracking this facility modification as FM-037. Once this system is installed, load characteristics will be evaluated to see if nuclear instrumentation can be added to the system.

NLW Detector, Cabling, and Connector repair: It is anticipated that the cable procurement will be the longest-lead item affecting an attempt at a repair. This repair is being tracked in the corrective action program (CA027).

Chambers: Two new spare fission chambers of identical design are available at the facility.

Cables: There is not sufficient supply of the engineered nuclear detector cable RSS-6-105/LE and jacket material at the facility to initiate a repair at this time. The cable must be one continuous segment from the detector to the preamp. The original manufacturer is Rockbestos Company, which became RSCC, which was purchased by Marmon Industrial Energy & Infrastructure. The manufacturer and distributors have been contacted, this engineered cable is manufactured only on an as-needed basis, requires an estimated 8-week lead time, with minimum purchase of 1000 ft. Procurement of this material will be sent to AFRRIs purchasing department as soon as quotes are finalized.

AFRRI - EN57730 Follow-up Report

Page 20 of 24 Connectors: The proper connectors have been identified and are being procured. The proper tools to install the connectors are available at the facility when the cabling and connectors arrive.

Conduit/Jacket material: Sufficient flexible conduit is available at the facility. The armored jacket material is not available at the facility but is being procured.

9. Summary The equipment failure during startup on 22 May 2025, resulted in a high-power SCRAM during startup in steady-state mode under two (2) control rod automatic control. During power escalation, high frequency noise induced in the log channel detector input causing the channel to fail to a fixed high frequency, failing to provide an accurate period signal to the automatic control system, which cause the two control rods to withdraw at the maximum rate. The reactor power level increased rapidly, until the reactor protection system high flux safety channels caused an automatic SCRAM, releasing control rods to safely shut down the reactor. Due to the rate of power increase, reactor power momentarily exceeded the state-state license limit of 1.1 MW for approximately 100 msec with a peak power level of 1.2 MW. All reactor protection systems functioned as designed. Neither the fuel temperature Safety Limit (1000 °C), nor the Limiting Safety System Setting (600 °C), were exceeded during the event. Maximum fuel temperature recorded was 36 °C. All radiation monitoring systems were operable during the event and no release of radioactive material was detected. No damage to the fuel occurred.

The event was reported to NRC Headquarters Operations Office in a timely manner (EN 57730).

Subsequent troubleshooting revealed that the source of the noise is related to AC-grounds. Noise in nuclear instrumentation signals, particularly the NLW has been an ongoing issue at AFRRI since before restart of the reactor in 2022, with several incremental improvements made since the initial restart in July 2022.

All current operators were briefed on the event with increased emphasis to be aware of these activities, indications, and failure modes during operation. The following short-term operational restrictions were put in place and emphasized during the training:

Minimize the use of 2-rod automatic for routine automatic startups until auto mode drive speeds can be tuned properly.

If indications of NLW noise become apparent during an automatic startup, immediately place the system in manual, and dispatch an operator to the bay to immediately correct the issue in accordance with Tech Spec 3.2.1.a.(2).

The following longer-term actions are being pursued:

Fission chamber, cabling, jacket, and connector replacement on NLW once materials can be procured, but no later than the upcoming annual maintenance outage scheduled for November/December 2025.

AFRRI - EN57730 Follow-up Report

Page 21 of 24 Transient rod drive motor modifications to use a stepper motor similar to other drives rather than AC circuits switching a DC motor to eliminate a primary source of noise.

Drive speed tuning for various modes of operation and number of drives being controlled in automatic to reduce the reactivity insertion rate in during automatic control.

Investigation of the condition of building AC ground network and modification of the electrical distribution system to improve grounding based on results of the investigation.

AFRRI - EN57730 Follow-up Report

Page 22 of 24

Appendix A:

Memorandum for Reactor Staff, Training, Temporary Restrictions, and Path Forward Following High Power Scram

AFRRI - EN57730 Follow-up Report

Page 23 of 24 Appendix B:

SCRAM 0027 Report Form

SCRAMS,AlarmsandAbnormalConditions 004 Revision2 17JUN2024

Page3of9

Figure1RxshutdownafterRun4,Exp3

Figure2Reactorparametersbeforeautomodeenabled

SCRAMS,AlarmsandAbnormalConditions 004 Revision2 17JUN2024

Page4of9

Figure3AutoModecontinuestopullrodspastcritical

Figure42secondspriortoSCRAM

SCRAMS,AlarmsandAbnormalConditions 004 Revision2 17JUN2024

Page5of9

Figure51secondpriortoSCRAM

Figure6MomentpriortoSCRAM

SCRAMS,AlarmsandAbnormalConditions 004 Revision2 17JUN2024

Page6of9

Figure7HistoryFileat10:55:25update(preSCRAM)

Figure8AtSCRAM

SCRAMS,AlarmsandAbnormalConditions 004 Revision2 17JUN2024

Page7of9

Figure9AtNP/NPPSCRAMsignal

Figure10SCRAMsandWarningscomein;SystemSCRAMSet.

SCRAMS,AlarmsandAbnormalConditions 004 Revision2 17JUN2024

Page8of9

Figure11MagnetPowerOff

Figure122secondsafterSCRAM

SCRAMS,AlarmsandAbnormalConditions 004 Revision2 17JUN2024

Page9of9

Figure135minutesafterSCRAM

AFRRI - EN57730 Follow-up Report

Page 24 of 24

Appendix C:

Corrective Maintenance Item CM 25-015 Troubleshooting sources of Noise in NLW Signal.

CorrectiveMaintenance 110 Revision0 18SEP2024

Page3of5

23MAY2025:

1128:Consoleon,airoff,magnetson-Logpowernormalat68e8 1129:ManipulatedB5IFE(drytubclosesttoNLWdrytube)-nochange 1129:ManipulatedC2IFE(normallysittingonKaptontape)-nochange 1131:Overcorelightson/off-nochange 1132:Jostledcorebyhand-slightchange~1/4decade,8e81.3e7 1133:Horizontal(alongcoremovementplane)movementoflogchannelconduit-nochange 1134:Tangentialmovementoflogchannelconduit-nochange 1136:Movingtransientroddrive(noair)-largespike,withslowdecreaseto2e7 1137:TRdriveup-largespike,signal34e7 1138:TRdriveup-largespikes,signal11.5e6 1139:TRdriveupto416units-largespike,signal7e61e7 1140:MovedTRdrivecable-quickdrop,slowdecayto5e7 1141:Movingtransientroddrivecablecausesnoise 1143:SlightadjustmenttoNLWcableattrunktocoreconnection-noisedroppedto~1e7 1144:Drivetransdrivedown-bigspike1e5 1145:TRdriveupto120units~6e7 1147:TRdrivedown-spike,stableat57e7 1147:Transfire-nochange 1148:Transupto100units-57e7 1149:Jostledcore~1/2decadedropinsignal 1150:Signaldownto~1e7 1152:Poollightson-spiketo1e3,slowdecayafter 1153:SWpoollightoff-drop,thenspikedto2.7e3 1154:SWpoollighton-dropto1e3 1156:NEpoollightoff-nochange 1156:SEpoollighton-1.21.3e3 1157:NEpoollighton-nochange 1157:SEpoollighton-downto~79e4 1158:NWlightoff-increaseto1.3e3 1159:NWlighton-switchingnoise 1159:SWlightoff-bigdrop,thenincreaseto22.5e7 1201:SElightoff-downto1.5e3 1202:touchedNLWcableatcoretotrunk(~4inmovement)-signaldropto2.3e4

CorrectiveMaintenance 110 Revision0 18SEP2024

Page4of5 1203:SWlighton-signalupto5e4 1203:SElighton-switchingnoise,thendropto~2.4e4 1204:movedNLWcableattrunktocore-dropto56e8 1206:OpenedNLWpreampcabinet;manipulatedground,signalcable,jacket,flexconduit-nochange 12071208:ManipulatedNLWflexconduitontrunk-nochange 1208:Poollightson/off-nochange 1209:AppliedTRair-nochange 1210:TRto180units-spikethenbackto8e8 1210:TRairoff,drivedown-spike,nochange58e8 1211:Rxbaylightson/off-nochange 1212:3157lightson/off-nochange 1213:PressPbdooropen-nochange 1214:pbdoorsopening,rxSCRAM-nochange 1216:pbdoorsopen-smallspikeonpbmotoroff 1217:movingcoretoregion1-smallspikesonmotoractuation 1220:playedwithcoremovementswitch-onlymotoractuationspikes 1221:enterregion1-smallspike 1222:atpos250,movedto264-nochange 1224:TouchedNLWcableatcoretrunk-smallspike 1224:TRup300units-nochange 1225:TRdrivedown-smallspike 1225:movedNLWcabletowardTRdrivecable-bigspike,butsettled 1226:movetoregion3 1229:Bayandpoollightson/off-nochange 1232:TRairon-nochange 1232:TRup-smallspike,nochange 1233:TRdown-spikethensettle 1233:Shimupto500-nochange 1234:Shimdropped-smallspikeondropandonmagnetrecouple 1235:Safeupto450-nochange 1236:Regupto500anddrop-nochange 1251:MovingcoretolookatNLWflexconduitrouting 1251:Largespikewhencorestoppedat750-returnedtonormal 1253:Coreenteredregion2largespike,signalremainhighat~68e6 1255:ManipulatedNLWflexconduit~3in-signaldroppedtonormal 1255:Continuedtomanipulateconduit-onelargespike,thennochange,allnormal 1303:Coreatposition250-nochange 1308:Coreatposition750-largespikethennormal 13081347:Continuedtomovecore-nochange 1347:ManipulatedNLWflexconduitbetweenuppercoreuprightclampandziptie-largespike,stayedat~2.5e5 1353:DisconnectedNLWsignalcableforoscilloscopetestingofsignal-nothingfound 1402:NoisestillpresentwhenNLWreconnected 1404:Poollightoff-signalbacktonormal 1409:Poollighton-signalto8e7 1426:Manipulatedcable-signaldroppedtonormal(~9e8)

CorrectiveMaintenance 110 Revision0 18SEP2024

Page5of5