Process Safety Indicator Frameworks

Abstract

Process safety performance indicators are used to judge whether major accident risk at oil and gas, refining and chemical facilities is under control, yet practitioners in Indonesia face several frameworks that differ in purpose and structure. This article compares API RP 754, IOGP Report 456, the CCPS metrics guide and the UK HSE's HSG254, and positions them against Indonesian requirements for performance monitoring under the SMK3 regulation and the oil and gas safety management system (SMKM). The study is a document-based comparison drawing on a web search conducted on 2 November 2026 of official sites of standards bodies, regulators, an investigation board and journals, using six comparison criteria. API RP 754 and IOGP 456 use a four-tier model with standardised event definitions for public reporting, the CCPS guide is voluntary and aligned with API, and HSG254 derives indicators from each site's risk control systems. Indonesian regulations require performance monitoring, but no standardised set of process safety event indicators was found in the documents that could be accessed. The scientific literature questions how sharply leading and lagging indicators can be separated, and evidence on the predictive value of indicators remains limited. The author concludes that the frameworks are better used together, with standardised event definitions for recording and benchmarking and site-specific, quality-tested barrier health indicators beneath them.

Keywords: process safety indicators, API RP 754, leading and lagging indicators, barrier management, SMKM

Introduction

The scholarly debate on process safety indicators has run for more than 15 years and is not settled. Hopkins (2009) separated two dimensions that are often conflated: personal safety versus process safety indicators, and lead versus lag indicators. In his account the first distinction is fairly clear, while the second is straightforward for personal safety but hard to pin down for process safety (Hopkins, 2009). A decade later, a review of the same literature found a wide variety of definitions, indicators that do not relate logically to current safety theories and models, and no single reliable indicator (Swuste et al., 2019).

The problem is not only academic. The explosions and fires at the BP Texas City refinery on 23 March 2005 killed 15 people and injured 180. The U.S. investigation board found that companies and the industry mostly tracked personal safety indicators rather than indicators able to reveal the risk of catastrophic failure, and it recommended that the American Petroleum Institute and the United Steelworkers union develop a standard for leading and lagging process safety indicators for nationwide public reporting and for use at individual facilities (U.S. Chemical Safety and Hazard Investigation Board [CSB], 2007). That recommendation led to the 2010 edition of API RP 754, whose fourth edition was published in August 2026 (American Petroleum Institute [API], 2026).

In Indonesia, Article 14 of Government Regulation No. 50 of 2012 on the Implementation of the Occupational Safety and Health Management System (SMK3) requires employers to monitor and evaluate OSH performance through examination, testing, measurement and/or internal SMK3 audits (Pemerintah Republik Indonesia, 2012). In oil and gas, Ministerial Decree No. 176.K/MG.01/MEM.M/2024 on guidelines for assessing implementation of the oil and gas safety management system includes an element on performance monitoring and measurement (Kementerian Energi dan Sumber Daya Mineral [KESDM], 2024). Both instruments impose a duty to monitor, but they do not in themselves answer which indicators should be monitored or which framework should be used.

The aim of this article is to compare the four most widely cited process safety indicator frameworks against explicit criteria and to position them against Indonesian requirements, so that organisations can choose or combine frameworks on reasoned grounds.

Methods

This is a document-based comparison of approaches, not a systematic review. The search was conducted on 2 November 2026 with a general web search engine, covering the official sites of API, AIChE/CCPS, IOGP, the UK HSE, the CSB, the Indonesian Ministry of Energy and Mineral Resources and peraturan.go.id, plus journal and university repository pages. Direct access to aiche.org, the TU Delft repository and Crossref was refused by the network used. Scopus, Web of Science and other bibliographic databases were not searched. Search terms included process safety performance indicators, API RP 754 tier, IOGP Report 456, CCPS process safety metrics, HSG254, leading and lagging indicators, indikator keselamatan proses and Sistem Manajemen Keselamatan Migas, with a source year range of 2006 to 2026.

Framework documents were included when issued by a standards body, regulator or industry association in their latest identifiable edition. Scientific articles were included when peer reviewed and concerned with the concept or quality of process safety indicators, and were read at the level of abstracts and publisher records. The full texts of API RP 754, IOGP 456 and the CCPS guide were not read; their content was taken from official publisher pages, press releases, CSB documents and search result excerpts, including excerpts from an unofficial copy of the CCPS guide that were treated as supporting information only.

The four frameworks were compared on six criteria: purpose and main users, indicator structure, basis for event definitions, comparability across facilities, link to barriers or risk control systems, and position relative to Indonesian regulation.

Results and Discussion

The Four-Tier Model

API RP 754 classifies indicators into four tiers. Tier 1 and Tier 2 indicators are regarded as suitable for public reporting by companies and the industry, while Tier 3 and Tier 4 indicators are intended for internal use (API, 2026). The third edition of 2021 adopted GHS criteria for threshold releases, made severity weighting of process safety events (PSEs) mandatory, added cause layers and extended data collection to non-petroleum chemical facilities (CSB, 2022). The fourth edition adjusts the direct cost thresholds for Tier 1 and Tier 2 for inflation and refines guidance on assessing fires, explosions and unignited releases (API, 2026).

IOGP Report 456 carries the same model into upstream operations. Its third edition was published in May 2023, incorporates the revisions of the third edition of API RP 754 and clarifies the classification of PSEs related to well operations (International Association of Oil & Gas Producers [IOGP], 2023), so one event definition can serve both upstream and downstream facilities.

The strength of the model is comparability. Standardised Tier 1 and Tier 2 definitions allow figures to be aggregated across companies, and API uses the framework as the basis of its site assessment programme and annual PSE survey (API, 2026). Its weaknesses are also on the official record. In 2012 the CSB judged the framework useful but noted that reporting of the leading Tier 3 and Tier 4 indicators was far more limited and ambiguous than that of the lagging Tier 1 and Tier 2 indicators, and that the indicators would not permit comparisons among sites, corporations or national averages (CSB, 2012). The recommendation was closed only in 2022, with a status of acceptable alternative action, after the third edition appeared (CSB, 2022).

CCPS Metrics Guidance

The CCPS metrics guide is addressed to organisations that wish to prevent major process safety events by evaluating trends within their process safety systems (Center for Chemical Process Safety [CCPS], 2022). It covers lagging, near miss and leading metrics, applies severity weighting only to differentiate Tier 1 PSEs, states that no CCPS publication contains mandates, and was designed to be consistent with the 2021 API edition (CCPS, 2022). In the author's interpretation, its value lies in scope, since chemical manufacturers outside petroleum association reporting can use it, while its force depends entirely on corporate commitment.

The HSG254 Approach

HSG254 was produced jointly by the HSE and the Chemical Industries Association and sets out a six-stage process for building a programme to monitor the performance of process safety risk controls, aimed mainly at senior managers and safety professionals (Health and Safety Executive [HSE], 2006). Leading indicators are defined as a routine, systematic check that key actions or activities are undertaken as intended, and lagging indicators show when a desired safety outcome has failed or has not been achieved. Both are derived from the risk control systems that matter for the major accident scenarios at the site (HSE, 2006).

The logic runs in the opposite direction from the four-tier model. API and IOGP start from events that are given thresholds and then counted, whereas HSG254 starts from scenarios and their controls and asks what evidence shows that each control is working. This is consistent with research that derives indicators from the status of preventive barriers on the left-hand side of a bowtie, as tested for the loss of cooling of a post-reformer in an ammonia plant (Schmitz et al., 2021). The price is comparability, because indicators designed for one site generally cannot be summed with those of another.

Indonesian Regulatory Framework

PP 50/2012 places monitoring and evaluation of OSH performance as the fourth of five stages of SMK3 implementation, with results reported to the employer for corrective action (Pemerintah Republik Indonesia, 2012). The provision applies across sectors and does not distinguish personal from process safety.

In oil and gas, the SMKM assessment guideline issued on 25 July 2024 contains an element on performance monitoring and measurement alongside elements such as risk management, asset and installation management, emergency response and incidents, and is intended to integrate process safety with OSH, asset integrity, environmental protection and security management (KESDM, 2024). Minister of Energy and Mineral Resources Regulation No. 32 of 2021 governs technical inspection, safety examination, risk analysis and remaining life extension of oil and gas installations and equipment (KESDM, 2021). In the author's interpretation, data from these inspection duties fit naturally into Tier 4 or HSG254-style leading indicators, such as on-time inspection of critical equipment.

In the texts and summaries available to the author, no PSE definition, release threshold or rate formula standardised by an Indonesian regulator was found, a finding to be read with caution because the SMKM guideline annex was not read in full.

Critiques in the Literature

Three scientific critiques apply to all the frameworks above. First, the leading and lagging distinction is less sharp than the guidance suggests. Hopkins (2009) showed that even the independent panel report that followed Texas City at times conflated the personal-process and lead-lag dimensions, and that one of its main findings was that lag indicators remain useful provided they are the right kind. Second, the relationship between indicators and accident processes is far from perfect, and the use of occupational injury rates as a proxy for process safety has been criticised as rooted in Heinrich's safety pyramid (Swuste et al., 2019). Third, indicator quality is rarely tested. Selvik et al. (2021) concluded that a safety performance indicator is fit to support barrier management decisions only when it satisfies all SMART criteria, and they proposed a shortened STAR version, illustrated with an indicator of dangerous fluid overfilling events motivated by the Texas City accident.

These critiques do not invalidate any framework, but they show that a low Tier 1 count or a high Tier 4 score is not yet evidence that major accident risk is controlled.

Comparative Synthesis

The summary below is the author's interpretation of the sources discussed above, organised by the six criteria.

• Purpose and users: public reporting and benchmarking for API RP 754 and IOGP 456, voluntary company use for CCPS, and site management for HSG254.
• Indicator structure: tiers from consequential events to management system performance in the first three, and leading and lagging pairs per risk control system in HSG254.
• Basis for definitions: standardised thresholds, consequences and costs in the four-tier model, site scenarios in HSG254.
• Comparability: high for Tier 1 and Tier 2, low for Tier 3, Tier 4 and HSG254 indicators.
• Link to barriers: weak for Tier 1 and Tier 2, which record failures already incurred, strongest for HSG254 and barrier status approaches.
• Position in Indonesia: none is specifically mandated, but all can serve the monitoring duties of PP 50/2012 and the SMKM.

The author's reading is that the most defensible choice is to divide the work between frameworks: Tier 1 and Tier 2 definitions from API RP 754 or IOGP 456 record events consistently for comparison with parent companies and partners, while indicators below them are designed from site scenarios and barriers using HSG254 logic and tested with criteria such as STAR. Because Tier 1 events are by definition rare at a single site, their annual trend carries limited information. This last point is the author's inference and has not been tested empirically in the sources reviewed.

Limitations

The full texts of API RP 754 (fourth edition), IOGP 456 (third edition), the CCPS guide (version 4.1) and the SMKM guideline annex were not read, so thresholds, rate formulas and example indicators are not discussed, and part of the CCPS content rests on excerpts from an unofficial copy. Scientific articles were read at abstract level. No peer-reviewed Indonesian study on process safety indicators was found, which does not mean none exists. HSG254, the 2007 CSB report and the Hopkins article are more than ten years old but underpin the frameworks now in force. The OECD framework and other countries' regulator reporting systems are not covered.

Conclusions and Recommendations

API RP 754 and IOGP 456 are strongest on event comparability, the CCPS guide extends the same logic to the chemical industry on a voluntary basis, and HSG254 is strongest on linking indicators to site risk controls. Indonesian regulation requires performance monitoring without standardising process safety event indicators in the documents that could be accessed, which leaves organisations both the freedom and the responsibility to choose. None of the frameworks has been shown to predict major accidents.

Practical recommendations for oil and gas, refining and chemical facilities:

• Adopt one Tier 1 and Tier 2 PSE definition, following API RP 754 or IOGP 456 according to the line of business, and record the edition used, because thresholds change between editions.
• Derive leading indicators from major accident scenarios and their critical barriers, for example function test results for safety devices, on-time inspection under Regulation No. 32 of 2021, and closure of HAZOP actions.
• Test each indicator against explicit quality criteria before using it for decisions.
• Report process safety indicators separately from occupational injury rates in SMK3 management reviews and SMKM assessments.

Further research should test the relationship between leading indicators and Tier 1 and Tier 2 events using Indonesian facility data, compare the reporting practices of upstream and downstream companies under the SMKM, and assess whether PSE severity weighting changes improvement priorities.

For organisations building or reviewing an indicator set, this comparison can be used alongside a process safety management review that traces the links between scenarios, barriers and indicators at their own facilities.

This article is general information, not legal or professional advice for any specific case.

References

1. American Petroleum Institute. (2026, August 6). API publishes fourth edition of Process Safety Performance Indicators standard [Press release]. https://www.api.org/news-policy-and-issues/news/2026/08/06/api-publishes-fourth-edition-process-safety-performance-indicators-standard
2. Center for Chemical Process Safety. (2022). Process safety metrics: Guide for leading and lagging indicators (Version 4.1). American Institute of Chemical Engineers. https://www.aiche.org/sites/default/files/docs/pages/ccps_process_safety_metrics_-_v4.1.pdf
3. Health and Safety Executive. (2006). Developing process safety indicators: A step-by-step guide for chemical and major hazard industries (HSG254). HSE Books. https://www.hse.gov.uk/pubns/books/hsg254.htm
4. Hopkins, A. (2009). Thinking about process safety indicators. Safety Science, 47(4), 460-465. https://doi.org/10.1016/j.ssci.2007.12.006
5. International Association of Oil & Gas Producers. (2023). Process safety: Recommended practice on key performance indicators (IOGP Report 456, 3rd ed.). https://www.iogp.org/bookstore/product/process-safety-recommended-practice-on-key-performance-indicators/
6. Kementerian Energi dan Sumber Daya Mineral. (2021). Peraturan Menteri Energi dan Sumber Daya Mineral Nomor 32 Tahun 2021 tentang Inspeksi Teknis dan Pemeriksaan Keselamatan Instalasi dan Peralatan pada Kegiatan Usaha Minyak dan Gas Bumi [Minister of Energy and Mineral Resources Regulation No. 32 of 2021 on technical inspection and safety examination of oil and gas installations and equipment]. https://jdih.esdm.go.id/storage/document/Permen%20ESDM%20No.%2032%20Tahun%202021.pdf
7. Kementerian Energi dan Sumber Daya Mineral. (2024). Keputusan Menteri Energi dan Sumber Daya Mineral Nomor 176.K/MG.01/MEM.M/2024 tentang Pedoman Penilaian Penerapan Sistem Manajemen Keselamatan Migas [Ministerial Decree No. 176.K/MG.01/MEM.M/2024 on guidelines for assessing implementation of the oil and gas safety management system]. https://migas.esdm.go.id/cms/uploads/regulasi/2024/b188cd316c00430a71b8442ccc3ca9fb.pdf
8. Pemerintah Republik Indonesia. (2012). Peraturan Pemerintah Nomor 50 Tahun 2012 tentang Penerapan Sistem Manajemen Keselamatan dan Kesehatan Kerja [Government Regulation No. 50 of 2012 on implementation of the occupational safety and health management system]. https://peraturan.go.id/files/pp50-2012bt.pdf
9. Schmitz, P., Swuste, P., Reniers, G., & van Nunen, K. (2021). Predicting major accidents in the process industry based on the barrier status at scenario level: A practical approach. Journal of Loss Prevention in the Process Industries, 71, 104519. https://doi.org/10.1016/j.jlp.2021.104519
10. Selvik, J. T., Bansal, S., & Abrahamsen, E. B. (2021). On the use of criteria based on the SMART acronym to assess quality of performance indicators for safety management in process industries. Journal of Loss Prevention in the Process Industries, 70, 104392. https://doi.org/10.1016/j.jlp.2021.104392
11. Swuste, P., van Nunen, K., Schmitz, P., & Reniers, G. (2019). Process safety indicators, how solid is the concept? Chemical Engineering Transactions, 77, 85-90. https://doi.org/10.3303/CET1977015
12. U.S. Chemical Safety and Hazard Investigation Board. (2007). Investigation report: Refinery explosion and fire, BP Texas City, Texas, March 23, 2005 (Report No. 2005-04-I-TX). https://www.csb.gov/assets/1/20/CSBFinalReportBP.pdf
13. U.S. Chemical Safety and Hazard Investigation Board. (2012). Board concludes American Petroleum Institute has made acceptable progress on recommendation to develop process safety indicators for onshore industries [News release]. https://www.csb.gov/board-concludes-american-petroleum-institute-has-made-acceptable-progress-on-recommendation-to-develop-process-safety-indicators-for-onshore-industries/
14. U.S. Chemical Safety and Hazard Investigation Board. (2022). Status change summary: Recommendation 2005-04-I-TX-R6a to the American Petroleum Institute. https://www.csb.gov/assets/recommendation/Status_Change_Summary_API_%28BP_TXC_R6a%29_C-AAA.pdf