Accelerator Reliability Workshop 2026 (ARW2026)

15 Nov 2026, 13:30 → 20 Nov 2026, 20:00 Asia/Tokyo

Hotel Aomori

Keitaro Kondo (QST)

The Accelerator Reliability Workshop (ARW) is a biennial international forum that brings together professionals from accelerator communities and related industries worldwide. It offers a unique opportunity to exchange experiences, ideas, and solutions aimed at improving the reliability of accelerator research and medical facilities.

ARW2026 will take place in Aomori City, Japan, from November 15 to 20, 2026. The workshop emphasizes interactive discussions and peer-to-peer engagement, fostering a collaborative environment for in-depth exploration of technical challenges related to accelerator systems and their operational availability.

Workshop Objectives:

• Share technical challenges and solutions that drive the reliability of accelerator facilities
• Promote cross-industry collaboration and future opportunities
• Encourage active dialogue across all organizational levels
• Explore the influence of cultural and human factors on reliability

Workshop Format:

Approximately 50% of the workshop time will be dedicated to discussion sessions. These sessions provide participants with opportunities to raise questions, share experiences, and receive feedback from peers who have faced similar challenges. This format ensures active involvement and helps build a strong collaborative network.

Topics Include:

• Reliability and performance optimization
• Engineering and design best practices
• Failure and reliability analysis methods
• Machine protection
• Maintenance strategies
• Integration of AI and machine learning
• Medical and commercial applications
• Future challenges and opportunities

ARW2026-Template_v01.pptx

ARW2026_poster.pdf

Sunday 15 November

Registration

Welcome Reception

Monday 16 November

LOC/IOC Welcome and Workshop Opening

10:00 Coffee Break

Panel Discussion

12:00 Lunch Break

Plenary Discussion Session

15:00 Coffee Break

The Cost of Unreliability: Failures, Consequences, and Recovery

Convener: Kathleen Genge (TRIUMF)

1 ISIS Target Station one, Recovery from a sticky position

At the heart of the ISIS Neutron and muon source are two target stations. Each target station uses a 25 metre long trolley system to transfer the Target Reflector and Moderator assembly from its operational position in the centre of the target station monolith, to a hot cell for maintenance. This approach reduces the duration of maintenance shutdowns and avoids the need to dismantle shielding and plant equipment to access critical components.
In July 2025, after over 40 years of operation, the Target Station one trolley system, weighing 400 tonnes, became jammed partway between the operating and maintenance positions, preventing operation of the Target Station.
This presentation describes the story of overcoming the engineering challenges of a recovery operation which had never been tried before in the history of ISIS. It will detail the rapid response to design, test and deploy tooling to remotely remove the obstruction, and will discuss some of the key challenges faced; being out of reach and visibility of the hot cell, limited crane coverage, narrow one-meter square aperture for importing all equipment into the cell, necessity to be controlled with remote manipulators.

Speaker: Mark Shaw (STFC - ISIS)

2 Recovering Maintainability Through Interface Reconstruction in the LIPAc RFQ Cooling System

The Linear IFMIF Prototype Accelerator (LIPAc) is jointly developed by Europe and Japan. The RFQ and its cooling system were procured by INFN-LNL, Italy, and delivered to Japan as an in-kind contribution. The cooling system is a complex subsystem consisting of six pumps with multiple temperature-control loops and is controlled by a hybrid Schneider-Siemens PLC architecture. Since commissioning in 2018, it has supported reliable RFQ operation through local control and EPICS-based supervisory control.
A major challenge arose when the company that designed and manufactured the cooling system ceased operations. As a result, detailed design information, including PLC ladder programs, and vendor support became unavailable. The integrated Japanese-European team operating LIPAc was therefore required to maintain a critical subsystem whose functionality and interfaces were only partially documented.
To address this issue, the joint team has collaborated with a Japanese engineering company experienced in accelerator control systems and EPICS integration. The existing system was investigated to reconstruct undocumented interfaces, including the boundaries between local control, supervisory control, and protection functions. The recovered interface definitions and system functions formed the basis of a modernization plan to replace the obsolete Schneider PLC with a maintainable Siemens platform. The project has now entered the detailed design phase.
This contribution presents the investigations and engineering studies carried out to reconstruct lost interface definitions from operational knowledge and remaining system assets. The challenges encountered, lessons learned, and their relevance to future accelerator projects requiring long-term support of internationally developed systems are discussed.

Speaker: Keitaro Kondo (QST)

3 Parallel operation and how we intend to resume operations after the fire.

At GSI, during the 2025 beam time, the accelerator complex achieved its highest availability in many years. Everything was geared toward realizing the First Beam Event for FAIR by the end of 2026. At the beginning of 2026, however, a fire occurred during a test run in the RF system. The entire RF system was destroyed, and many devices were severely affected by smoke and firefighting water. Rebuilding will take several years. Interim solutions to enable temporary scientific operation were planned and are to be implemented.

Speaker: Markus Vossberg (GSI Germany)

Tuesday 17 November

Beyond Firefighting: Implemented Lessons Learned for Improved Reliability

Convener: Martin Carroll (European Spallation Source)

4 From Reactive Operations to Long-Term Reliability at the Linac Coherent Light Source

The Linac Coherent Light Source facility comprises a normal-conducting linear accelerator and a newly commissioned superconducting linear accelerator, both capable of multiplexed electron beam delivery to two independent undulator lines. Maintaining stable and reproducible accelerator performance in this environment has often required reactive troubleshooting to address recurring operational issues.

To improve long-term reliability and operational efficiency, focused efforts have been undertaken to identify the root causes of persistent performance variability across both accelerator systems. This presentation highlights the transition from reactive “firefighting” toward a more sustainable and resilient operational model through targeted enhancements in hardware reliability, automation, feedback systems, and operational procedures. Lessons learned from these implementations will be discussed, along with their impact on accelerator stability, reproducibility, and machine availability.

Speaker: Christopher Zimmer (SLAC National Accelerator Laboratory)

5 Does process and knowledge management improve the availability of accelerators through the systematic application of lessons learnt?

The availability of accelerators is a critical success factor for scientific research institutions.
Reliable operational processes and their continuous optimization are key drivers. Both aspects are an integral part of a comprehensive quality management (QM).
This presentation demonstrates that systematic process management, along with the implementation of operational lessons learned, achieve the opposite: it can stabilize and enhance the availability of facilities.
A show-case, illustrates how traditional firefighting - the ad-hoc resolution of failures without in-depth root cause analysis - may leads to recurring problems. This approach is inefficient and undermines the availability of work processes and technical systems. In contrast, the analysis of the failure by the quality department, along with the institutionalization of findings through the PDCA cycle and process modeling, demonstrates how robust solutions can be developed.
This highlights that the underlying - often organizational - root causes of failures can only be uncovered and resolved through a systematic, methodical investigation.
QM should be understood not as an obstacle but as an enabler for research: By optimizing work processes and establishing quality methods, not only the availability of accelerators can be increased, but the freedom of research could also be strengthened - as reliable processes allow more time for actual research work rather than waiting for repeated repairs.
The results underscore that lessons learned and a continuous improvement process of internal operational workflows are key components for the reliability of scientific infrastructures.

Speaker: Mr Siegfried Koepke (DESY Deutsches Elektronen Synchrotron)

6 From Firefighting to Sustainment - An Integrated Framework for Accelerator Reliability

Mature accelerator facilities are often highly effective at restoring operation after failures, yet repeated reliance on experienced personnel and rapid technical fixes can unintentionally reinforce a reactive “firefighting” culture. At the 54-year-old Los Alamos Neutron Science Center (LANSCE), operational experience has motivated a broader approach: systematically converting failures, maintenance experience, and lessons learned into actions that improve long-term reliability.
This presentation describes LANSCE’s evolving integrated sustainment framework, which connects Asset Management, Risk Management, Conduct of Engineering, Conduct of Operations, Training & Qualification, and Data Collection & Analysis through a common accelerator system hierarchy.
Operational events and equipment-performance data help identify recurring problems and risks, prioritize corrective actions and upgrades, strengthen engineering and operational practices, improve documentation and training, and evaluate whether implemented actions are producing the intended results.
Current implementation efforts illustrate how LANSCE is moving from predominantly reactive responses toward repeatable, risk-informed processes that help balance immediate beam-delivery needs with longer-term reliability improvements.

Speaker: Martin Pieck (Los Alamos National Laboratory)

10:00 Coffee Break

Breakout Session I

Breakout Session II

12:00 Lunch Break

Poster session

15:00 Coffee Break

Diligent Incident and Fault Tracking and Effective Use of Data

Convener: Lukas Felsberger (CERN)

7 From Faults to Availability: A Decade of Operational Experience with CERN's Accelerator Fault Tracking System

More than a decade after its deployment, CERN's Accelerator Fault Tracking (AFT) system has become the primary reference for availability data across the CERN accelerator complex. Its success relies on close collaboration between Accelerator Operators (OP), Equipment Groups, and the Reliability and Availability Working Group (RAWG).
Extending AFT from the LHC to the full Injector Complex and Experimental Areas introduced new challenges. In fast-cycling machines, many faults are too brief for reliable manual recording, requiring automated correlation of interlock signals and operational logs with fault records. The solution developed at CERN is presented as a model for facilities facing similar operational constraints.
Also, maintaining data quality over the long term requires continuous attention. This contribution describes the regular review of system categories and expert notification groups, together with a structured user feedback process that has enabled iterative improvements and sustained community engagement.
AFT data underpin operational reporting at multiple levels, including weekly fault review meetings, high-level committees, and annual workshops. They also support consolidation decisions by providing the operational history needed to prioritize hardware renewal.
Finally, the CERN RAWG derives equipment failure and demand rates from the AFT dataset to support availability modeling for future accelerators. The ongoing FCC availability study illustrates how a high-quality fault history accumulated over years of operation becomes a key input to next-generation machine design.

Speaker: Gian Piero Di Giovanni (CERN)

8 Reliability and Availability at HIPA: A User-Driven Evolution (2024-2026)

Following discussions at ARW2024 and continued exchanges with HIPA users at PSI, we modernized the reliability and availability reporting framework of the PSI High Intensity Proton Accelerator (HIPA) to better reflect experiment‑relevant performance and improve trust in the published statistics. Traditionally, availability was reported only for beam currents above 1000 µA, and outages were logged in detail only when exceeding five minutes.
User feedback highlighted strongly differing sensitivities: some experiments depend mainly on delivered charge, others require sustained high current, and many are particularly impacted by interruptions longer than ~30 minutes. In response, starting in 2024 we refined outage classification by introducing a dedicated category for interruptions longer than 30 minutes to distinguish high‑impact events from the broader set of >5‑minute outages. Beginning in 2025, we expanded reporting beyond fixed thresholds by collecting and analysing the full distribution of trip and outage durations, enabling advanced visualizations (e.g., Pareto analyses) that reveal frequency and operational relevance across timescales.
For 2026, the framework converges toward standardized, user‑aligned definitions and reporting logic:
(i) two availability figures at MHC1 will be published for low current (1000 µA) and high current (1450 µA),
(ii) the reporting “day” follows the operational scheduling boundary (08:00–08:00), and (iii) a single documentation threshold of 15 minutes is used to reduce reporting overhead while preserving the dominant downtime and cause picture.
Users also pointed out that rapid sequences of consecutive trips, with time‑to‑failure intervals below ~30 seconds, are effectively indistinguishable to their instrumentation and behave as a single extended interruption. Starting in 2026, such clusters are aggregated and reported as one outage, and short intervals of nominal operation between them are no longer counted as productive uptime.
This presentation will introduce the updated definitions, reporting logic, and latest operational results, and will report progress toward automated cause classification of short interruptions using interlock system information and enhanced logic.

Speaker: Mr Antonio Barchetti (Paul Scherrer Institut)

9 Fault Analysis of CSNS I Linac DTL RF System and Its Operational Implications for CSNS II Upgrade

The CSNS I linac delivers 80 MeV H⁻ beam as injector for the Rapid Cycling Synchrotron. Its RF system consists of RFQ, bunchers and four klystron based DTL power sources. The digital LLRF control system achieves cavity field stability of 1 % in amplitude and ±1° in phase. Based on operational fault statistics covering the beam production period from Oct. 2025 to Jun. 2026, total RF related beam downtime reached 104.1 hours, among which DTL1-4 klystron power source failures contributed 26.3 hours, accounting for 25.3 % of linac RF induced outages. DTL3 exhibited the highest fault frequency, dominated by abrupt reflected power excursions that triggered downstream RCS beam loss interlocks. DTL4 frequently suffered unannounced high voltage trips originating from electromagnetic interference to its monitoring unit. Additional recurrent operational issues include LLRF hardware lock ups, parameter configuration loss following software restarts, and spurious interlocks caused by marginal sensor threshold settings. This paper presents observed fault signatures, on shift mitigation procedures and root cause analysis. Practical improvements are proposed, including enhanced alarm coverage and revised operational SOPs. Lessons learned from the room temperature DTL RF operation are summarized, providing practical references for operation oriented design and operator training of the next generation linac RF system for the CSNS II upgrade.

Speaker: Meifei Liu (Institute of High Energy Physics, Chinese Academy of Sciences)

Banquet Dinner

Wednesday 18 November

Innovative Systems and Robotics for Reliable Accelerators

Convener: Mario Di Castro (CERN)

10 Anomaly detection trial at J-PARC accelerator

More than twenty years have passed since the construction of J-PARC began. In recent years, troubles caused by aging equipment have become increasingly frequent. To improve an operational availability and enhancing user convenience, it is essential to adress the aging issue. This presentation reports on the development status of an anomaly detection system that applies machine learning. The system attempts to detect abnormal signals using acoustic, vibration, and odor sensors.

Speaker: Kazami Yamamoto (J-PARC Center)

11 How Much Testing Is Enough? Endurance Qualification of the SPES Remote Handling Framework during Cold Commissioning

At the SPES radioactive ion beam facility (INFN Legnaro National Laboratories), the Target Ion Source unit requires periodic replacement under radiological conditions precluding personnel presence. The operation is entrusted to a dedicated Remote Handling framework: an automated guided vehicle for transfer within the target area and a cartesian manipulator serving a shielded decay rack of 54 positions. Both are designed for fault tolerance, with redundant actuation and sensing, under a distributed control architecture whose safety functions are certified up to SIL 3.
Commissioning such a framework raises the question of how much endurance testing qualification requires. The approach presented derives a reliability target per mission from the foreseen operating profile, fewer than ten replacements per year with recovery delayed by radioactive decay, and converts it into the number of test missions required at stated confidence. During cold commissioning, the campaign dimensioned on this basis is structured on two levels: repeated transfers across all storage positions, providing the statistical basis, and fewer complete transfer missions, exercising the entire replacement chain and its interfaces. Given the infrequent demand, deliberate standstill periods expose dormant failures and size the periodic exercise programme foreseen for operation. Test data are archived in the open source EPICS platform.
Because the campaign compresses duty cycles but not calendar time, the parametric signatures recorded form the pre-irradiation baseline against which the effects of accumulated dose will be assessed in operation.

Speaker: Giordano Lilli (INFN-LNL)

12 Wireless IoT Rad-Tol for Improved Reliability and Availability of Particle Accelerator Infrastructure

Particle accelerator facilities operate in environments where radiation, limited accessibility, and long intervention times can affect equipment reliability and machine availability. Monitoring environmental and operational conditions is therefore essential for fault diagnosis and maintenance.

At CERN, a radiation-tolerant wireless IoT platform was developed initially around the Battery Radiation Monitor (BatMon), providing distributed radiation measurements to correlate equipment failures with the local radiation environment. The platform has evolved into a multi-purpose sensing infrastructure supporting temperature, humidity, water-leak, vacuum monitoring, and lately monitor accelerator communication infrastructure, while reducing cabling and installation requirements.

The technology is being explored beyond CERN through collaboration with J-PARC, where BatMon devices have been deployed for radiation monitoring and reliability studies of RF equipment, including Finemet amplifiers.

This evolution demonstrates the potential of radiation-tolerant wireless sensing to increase infrastructure observability, improve fault diagnosis, simplify instrumentation deployment, reduce intervention time, and ultimately contribute to higher accelerator availability.

Speaker: Salvatore Danzeca (CERN)

10:00 Coffee Break

Simulation versus Test Benches for Accelerator Reliability

Convener: Dr Ken Baggett (Jefferson Lab)

13 Development and Validation of a Nondestructive Capacitive Pick-Up type Bunch Shape Monitor

Monitoring the longitudinal bunch shape is important for understanding beam dynamics and maintaining beam quality in accelerator facilities. Conventional bunch shape monitors are often interceptive, making continuous measurements difficult during beam operation. Nondestructive approaches based on residual gas have also been developed to reduce beam interaction; however, these systems are generally more complex and require greater effort for installation and maintenance than capacitive pick-up monitors.
We developed a nondestructive Capacitive Pick-Up type Bunch Shape Monitor (CPU-BSM) for nonrelativistic ion beams. The CPU-BSM detects electrical signals induced on a capacitive pick-up electrode without inserting material into the beam path, enabling online bunch-shape monitoring with minimal interaction with the beam and offering advantages in terms of maintenance. For nonrelativistic beams, the longitudinal electric-field spread causes the measured signal to appear broader than the actual bunch. To correct this effect, a reconstruction algorithm was developed to recover the longitudinal bunch profile from the measured signal.
The monitor geometry and signal transmission structure were designed and optimized using electromagnetic simulations. Following fabrication, a dedicated coaxial test bench was developed to reproduce beam-like electric-field signals and verify the response of the fabricated devices before beam installation. The test bench also allows manufacturing and assembly issues in internal structures, which are difficult to inspect after fabrication, to be identified. The performance of the CPU-BSM was subsequently validated with approximately 25 MeV proton beams using a fast Faraday cup as a reference.
This presentation introduces the CPU-BSM and its development process, including electromagnetic design, bench testing, and beam validation.

Speaker: Donghyun Kwak (Korea Atomic Energy Research Institute (KAERI))

14 AvailSim4: A General-Purpose Simulation Tool for Accelerator Availability

Ensuring accelerator availability requires modelling of the interactions between component failures, repair processes, system dependencies and operational processes. AvailSim4 is an open-source, general-purpose, discrete-event Monte Carlo tool providing a simulation foundation for addressing these challenges across a range of accelerator projects and technologies. Its flexible architecture has supported studies ranging from contained systems to integrated accelerator complexes comprising thousands of components. Previous applications include reliability studies for MYRRHA and quench-detection systems in the LHC. For CERN’s proposed Future Circular Collider, an adapted AvailSim4 implementation provides the simulation engine for RAMI Lab, a dedicated environment developed to connect detailed system behaviour to accelerator performance. This presentation will showcase the adaptability of AvailSim4 across a spectrum of accelerator projects, illustrating how its simulation capabilities can support reliability-informed engineering.

Speaker: Jack Heron (CERN)

15 Safety Test Benches for ITER TBM Systems at QST Rokkasho: Experimental Validation of Safety Assessments

The Water-Cooled Ceramic Breeder (WCCB) Test Blanket System (TBS), a fusion blanket test module for ITER, employs high-temperature, high-pressure water cooling and requires safety assessments based on numerical simulations of thermal-hydraulic and accident phenomena. Because the accuracy of such assessments depends on modeling assumptions and input data, experimental validation is essential to establish confidence in safety analyses. This need to validate simulation tools against representative experimental data is a common issue in both fusion safety studies and accelerator reliability assessments.
To support validation of simulation tools used in ITER safety assessments, QST Rokkasho has developed a series of dedicated safety test benches that reproduce key phenomena under representative operating conditions. These facilities provide benchmark data for assessing analytical models and reducing uncertainties in safety evaluations.
Particular emphasis has been placed on two facilities that directly support simulation validation. A High Heat Flux Test Facility has been used to evaluate heat removal performance and thermal response of first-wall components under ITER-relevant heat loads. Experimental temperature and cooling-performance data are compared with analytical predictions to confirm thermal design margins and improve confidence in thermal-hydraulic models.
An In-Box Water Eruption and Leakage Test Facility has been developed to investigate loss-of-coolant accident (LOCA) scenarios inside the TBM box. Measurements of coolant release behavior and pressure transients provide benchmark data for validation of accident analysis methodologies and contribute to improved understanding of accident progression mechanisms.
In addition, a Flow Assisted Corrosion Test Loop is being operated to obtain material degradation data for future model development and long-term reliability assessments, while a Beryllium-Steam Reaction Test Facility provides experimental information on hydrogen generation and reaction heat associated with beryllium-water interactions.
This paper presents the objectives and capabilities of these safety test benches, with particular focus on the comparison between simulation results and experimental observations obtained in high heat flux and LOCA studies. The experience demonstrates how dedicated test facilities can improve model credibility, reduce uncertainties, and strengthen confidence in the safety and reliability of fusion blanket systems.

Speaker: Takanori Hirose (QST)

12:00 Lunch Break

Technical Tour

Thursday 19 November

Superconducting Systems

Convener: Charles Peters (ORNL-SNS)

16 Preparing the RHIC helium plant for future EIC operations

The Electron-Ion Collider (EIC) will require expanded cryogenic capacity, with the existing RHIC 4.5 K helium plant providing cooling for superconducting magnets, SRF systems, and the new detector magnet, as well as supplying and processing return helium for three new 2 K satellite plants around the EIC ring. Built in the early 1980s and refurbished for RHIC in the 1990s, the plant has operated reliably for 25 years but will face significantly higher loads: power rising from 7 MW to 9.5 MW and circulation increasing from 2,000 g/s to 3,000 g/s.

To meet these demands, five major end‑of‑life replacement and optimization projects are planned. These include replacement of Rotoflow turbo‑expander fluid control units, renewal of large intercooler and aftercooler heat exchangers, modernization of plant controls with new PLC-based systems, replacement of cooling tower #7, and upgrades to cold‑end heat exchangers and the final gas‑bearing turbo‑expander. Completion by 2030 will support commissioning of the first 2 K satellite plant. Reducing the plant’s helium leak rate—now elevated after the final years of RHIC operations—is also a key objective, driven by both operational reliability and rising helium costs.

This work is supported by Brookhaven Science Associates, LLC under Contract No. DE-SC0012704 with the U.S. Department of Energy.

Speaker: Russell Feder (Brookhaven National Lab)

17 Superconducting systems reliability at the Spallation Neutron Source

The Oak Ridge National Laboratory (ORNL) Spallation Neutron Source (SNS) linear accelerator (LINAC) operates with superconducting radio frequency (SRF) cavities maintained at 1.8 K. The cavities are cooled by a continuously running liquid helium refrigeration system. The cryogenic plant has supported uninterrupted operation for nearly two decades, enabling stable cavity performance and reliable accelerator availability. In this presentation, I will discuss our experience operating at such low temperatures with high reliability, how past and present challenges have been tackled, and how we are planning and preparing for higher power operation following the Proton Power Upgrade (PPU).

Speaker: Paolo Pizzol (Oak Ridge National Laboratory)

18 Efforts and Challenges Achieving and Maintaining High-Reliability SRF Systems at CEBAF

At Jefferson Lab’s (JLab) Continuous Electron Beam Accelerator Facility (CEBAF), superconducting RF (SRF) systems directly govern machine availability and energy reach. Sustaining high-reliability operations requires balancing high accelerating gradients against field emission, cryogenic heat load, trip rate, and trip recovery time. JLab has implemented strategic programs to maintain and improve SRF system performance, including cryomodule refurbishments, field emission amelioration, and low-level RF controls upgrades. This talk will focus on how past poor SRF system performance was overcome, and ongoing efforts to ensure long-term, reliable high-gradient operation.

Speaker: Daniel Moser (Jefferson Lab)

10:00 Coffee Break

Invited Session

12:00 Lunch Break

Improving Reliability Through Software

Convener: Dr Laurent NADOLSKI (Synchrotron SOLEIL)

15:00 Coffee Break

Breakout Session IV

Breakout Session Ill

Friday 20 November

Integrated Reliability across accelerator sub-systems – Importance of robust interface definition

Convener: Duane Newhart (Facility for Rare Isotope Beams (FRIB/MSU) - JLab)

19 The SNS Electron Catcher: Lessons on Reliability Impacts Resulting from Interface Definitions

At the Spallation Neutron Source (SNS), the proton accumulator ring concentrates technical risk in the injection region, where the electron catcher (e-catcher) removes secondary electrons generated during H⁻ stripping to protect nearby components and maintain stable operation. Following Proton Power Upgrade (PPU) modifications, the e-catcher failed on first beam operation due to insufficient cooling capacity. Other shortcomings of the design resulted from poor interface definition and technical review practices. This incident demonstrates how small specification and unit-consistency error can propagate through design reviews and lead to installation and operational issues. The work summarizes the failure mechanism, diagnostic process, and the need for more rigorous interface definition and technical-review practices.

Speaker: Melissa Harvey (Oak Ridge National Laboratory - Spallation Neutron Source)

20 Efforts at the SPring-8 campus to improve the operational reliability of the integrated accelerator complex

The accelerator facility in Harima consists of the CW ring-type light source “SPring-8,” which has been in operation for more than 30 years, and the relatively new X-ray free-electron laser (XFEL) “SACLA.” The aging SPring-8 is scheduled to shut down in the summer of 2027 and undergo a massive upgrade involving a complete replacement of its accelerators, transforming it into SPring-8-II.
Aiming for sustainable facility operations, we have been making efforts to optimize the entire accelerator complex to timely address aging issues. This presentation will first outline our long-term strategy for the sustainable facility, and then introduce two initiatives across the subsystems of the accelerator complex.
The first topic is the countermeasure to the aging beam injector dedicated for the SPring-8 storage ring. We decommissioned this injector and instead built the system utilizing the XFEL linear accelerator on a time-sharing manner to inject the beam into the ring.
The second topic concerns the countermeasure to enhance the reliability of the XFEL electron gun system, which lacks a backup unit, to fit the CW operations by SPring-8/SPring-8-II using a top-up beam injection.

Speaker: Hitoshi Tanaka (RIKEN SPring-8 Center)

21 The Interface Is a System: Managing Boundaries for Reliable Integration

Modern accelerator facilities are increasingly developed through distributed collaborations involving multiple teams, institutions, and external suppliers. While individual subsystems are often designed to high technical standards, many integration issues originate at the interfaces between them rather than within the subsystems themselves.
Ambiguous ownership, incomplete interface definitions, inconsistent engineering practices, and organizational barriers can significantly affect reliability, maintainability, and commissioning performance.
This contribution examines common organizational and technical challenges encountered during subsystem integration, including unclear scope boundaries, missing interface requirements, lack of standardized assembly and positioning guidelines, inconsistent component selection, and knowledge loss caused by staff turnover or outsourced development. A recurring challenge is that interface ownership is frequently avoided because it introduces additional responsibility, cost, coordination effort, and perceived risk, while providing limited recognition to the teams involved.
Drawing on practical engineering experience, this poster proposes a lightweight framework for improving interface management throughout the project lifecycle.
The framework emphasizes early definition of interface requirements, shared ownership between interfacing teams, Interface Control Documents (ICDs), responsibility matrices (RACI), configuration and change management, standardized engineering practices, and interface-specific verification before system integration. The objective is not to increase process overhead, but to reduce ambiguity and establish clear accountability where subsystem boundaries meet.
The poster  concludes with a set of practical recommendations that organizations can adopt to improve communication, facilitate collaboration across organizational boundaries, and increase overall system reliability. The central message is that subsystem interfaces should be treated as engineered products in their own right, requiring shared definition, shared accountability, and coordinated verification to achieve reliable operation throughout the facility lifecycle.

Speaker: Rafael Xavier da Silveira (European Spallation Source)

10:00 Coffee Break

Summary and Close Out

12:00 Lunch Break

Excursion