Insight: Validation of a hydrogen network RTO application for decisio

About the Authors and the Research

This paper was co-authored by Anibal Galan, C. Prada, and G. Gutiérrez, published in the IFAC proceedings (IFACOL) in 2018 and indexed in arXiv. The lead researcher, C. Prada, holds an h-index of 21 with a cumulative citation count of 1,634—a strong indicator of sustained academic influence in the fields of process control and industrial optimisation. The paper itself has been cited 6 times since publication, with its methodology being applied across refinery and process engineering contexts.

The research setting is a real refinery hydrogen network—a prototypically complex industrial system subject to fluctuating hydrogen demand, variable hydrocarbon (HC) loads, and equipment uncertainties. The team validated the RTO's utility as a decision-support tool for operators by comparing RTO-calculated optimal values against human-generated reconciled values (REC) across two distinct operational scenarios: a high hydrogen demand period (Period 1) and a low hydrogen demand period (Period 2).

Three Core Findings: When Systematic Optimisation Outperforms Human Coordination

The paper's fundamental contribution lies in its rigorous, data-driven validation of an RTO system's decision-support effectiveness using real industrial data—and in surfacing the systematic conditions under which human judgment falls short.

Finding 1: RTO Outperforms Human Coordination in Both Scenarios, but the Gap Is Scenario-Dependent

Across both the high-demand and low-demand periods, the RTO consistently produced better solutions than human operator reconciliation. However, the performance gap between RTO and REC was significantly larger during the low-demand period. This counterintuitive finding carries direct implications for BCM: organisations tend to assume that low-pressure operational periods are "safe," when in fact they may be when decision quality deteriorates most without detection.

Finding 2: Hydrogen Purification Membrane Utilisation Is the Key Differentiator

The study identifies the more effective use of hydrogen purification membranes by the RTO—relative to operator decisions—as the primary driver of the performance gap in low-demand scenarios. Managing membrane deployment requires simultaneous consideration of multiple interdependent network variables, a task that exceeds the cognitive bandwidth of real-time human decision-making. This is precisely the domain where Model Predictive Control (MPC) architectures excel, systematically searching for global optima that local human reasoning cannot reliably reach.

Finding 3: RTO Delivers High-Value Ancillary Decision Information

Beyond its primary optimisation outputs, the RTO system automatically provided operators with additional actionable intelligence: optimal HC load configurations, minimum gas purge recommendations, and optimal hydrogen production levels. This multi-dimensional decision support—delivered as a by-product of the core optimisation process—mirrors what a well-designed BCM framework should offer: not just a recovery time target, but a structured set of decision aids that remain active and useful throughout a disruption event.

Implications for Taiwan's BCM Practice: The Decision-Support Validation Gap

The central BCM implication of this paper is that when enterprises operate under uncertainty, human decision-making carries a systematic optimisation deficit—one that is harder to detect, and potentially larger, during non-peak operational states.

Mapping this to ISO 22301 Business Continuity Management practice, Winners Consulting Services Co. Ltd. identifies three structural blind spots prevalent among Taiwanese enterprises:

Blind Spot 1: Static RTO/RPO Targets Without Dynamic Validation. Most Taiwanese enterprises establish Recovery Time Objectives (RTO) and Recovery Point Objectives (RPO) in their Business Continuity Plans (BCP) based on initial Business Impact Analysis (BIA) estimates. Without a systematic mechanism to compare actual decision outcomes against optimised benchmarks—as demonstrated in this paper—BCP targets can silently drift from operational reality.

Blind Spot 2: ISO 22301 Continuous Improvement Clauses Remain Procedural. ISO 22301 Clause 10 mandates a genuine continuous improvement mechanism. However, most organisations fulfil this requirement through annual document reviews rather than through scenario-based, quantitative gap analysis comparing human decisions against systematic optima.

Blind Spot 3: Non-Critical Scenario Resilience Is Systematically Overestimated. BCM exercises in Taiwan are predominantly designed around major disaster scenarios. This paper suggests that moderate or low-pressure disruption scenarios—where decision quality tends to degrade silently—deserve equal if not greater analytical attention in BCM programme design.

How Winners Consulting Services Co. Ltd. Supports Taiwanese Enterprises

Winners Consulting Services Co. Ltd. (積穗科研股份有限公司) helps Taiwanese enterprises establish ISO 22301-compliant BCM frameworks, calibrate RTO and RPO targets, conduct Business Impact Analyses (BIA), and design crisis management exercises. Based on the findings of this research, we recommend the following three-phase action plan with a 7–12 month implementation timeline:

1. Months 1–3: BCM Decision Mechanism Diagnostic. Apply the paper's comparative methodology to map all critical decision nodes within the existing BCP, distinguishing between human-dependent and system-assisted decisions. Establish quantitative decision quality baselines and identify gaps relative to ISO 22301 BIA requirements.
2. Months 4–8: Scenario-Based Validation Framework Design and Implementation. Design BCM exercises covering both high-pressure and low-pressure disruption scenarios, mirroring the paper's dual-period validation structure. Integrate decision-support tools—including meta-heuristic algorithm-assisted resource scheduling—and establish a recurring mechanism to compare actual exercise decisions against optimised response paths.
3. Months 9–12: Continuous Improvement Mechanism and ISO 22301 Certification Readiness. Update BCP documentation to reflect validated RTO/RPO targets. Operationalise ISO 22301 Clause 9 (performance evaluation) and Clause 10 (continual improvement) through quantitative decision quality review reports, preparing the organisation for external certification audit.

Frequently Asked Questions

How does the RTO validation methodology in this paper apply to enterprise BCM frameworks?

The paper's core methodological contribution—systematically comparing actual operator decisions against RTO-optimised benchmarks—translates directly into BCM practice as a "decision quality gap analysis." Enterprises can apply this approach by comparing actual BCP exercise outcomes against pre-defined optimal response paths, quantifying the gap as a KPI. This is particularly valuable for manufacturing and energy sector organisations where operational decisions under uncertainty have significant continuity implications. The finding that gaps are larger in low-demand scenarios specifically suggests that BCM exercises should not exclusively focus on worst-case scenarios, but should include moderate and low-disruption scenarios to capture the full decision quality profile across operational states.

What are the most common compliance challenges Taiwanese enterprises face when implementing ISO 22301?

The three most frequently encountered challenges are: (1) BIA data quality—departments tend to provide conservative or incomplete RTO/RPO estimates, resulting in BCP targets that do not accurately reflect operational reality; (2) cross-functional integration gaps—BCM responsibility in many Taiwanese enterprises remains siloed within IT or information security functions, with insufficient integration of operations, supply chain, and finance; and (3) performative continuous improvement—ISO 22301 Clause 10 requires substantive ongoing improvement, but organisations frequently fulfil this with annual document updates rather than the quantitative scenario analysis that would identify genuine decision quality deficits. Addressing all three requires a structured governance model supported by experienced BCM advisors.

What are the core requirements of ISO 22301, and what does a realistic implementation timeline look like for Taiwanese enterprises?

ISO 22301 covers ten clauses spanning context analysis (Clause 4), leadership commitment (Clause 5), planning (Clause 6), support resources (Clause 7), operational execution including BIA and BCP design (Clause 8), performance evaluation (Clause 9), and continual improvement (Clause 10). A realistic timeline for mid-sized Taiwanese enterprises (500–2,000 employees) is 7–12 months: the first 3 months address gap analysis and leadership alignment; months 4–6 cover BIA, risk assessment, and BCP drafting; months 7–9 focus on exercises, testing, and document refinement; months 10–12 complete internal audit, management review, and external certification application. Typically, 1–2 dedicated internal BCM coordinators working alongside external consultants is the minimum viable resource model.

How should Taiwanese enterprises evaluate the cost and expected return on investment of BCM implementation?

For mid-sized Taiwanese enterprises, the typical investment in ISO 22301-compliant BCM implementation includes external consulting fees of approximately NTD 1.5–3 million over a 7–12 month engagement, plus certification audit fees of NTD 150,000–250,000. Quantifiable returns include a 30–50% reduction in mean recovery time for significant disruption events, reduced financial losses during operational interruptions, and measurable improvements in supply chain and customer trust metrics. Non-quantifiable returns include enhanced eligibility for government procurement contracts, stronger positioning in international supply chains, and improved insurance negotiation leverage. Winners Consulting recommends beginning with a complimentary BCM diagnostic to precisely identify gaps and prioritise investment, ensuring every resource committed produces measurable resilience improvement.

Why should Taiwanese enterprises choose Winners Consulting Services Co. Ltd. for BCM advisory?

Winners Consulting Services Co. Ltd. (積穗科研股份有限公司) brings a research-grounded, industry-specific approach to BCM consulting in Taiwan. Our advisors maintain active engagement with international academic research and regulatory developments—including CISA's Operational Technology security guidance and IFAC process control research—ensuring recommendations reflect current global best practice. Our end-to-end service capability covers BIA design, BCP development, RTO/RPO calibration, crisis management exercise facilitation, and ISO 22301 certification audit preparation. We serve enterprises across manufacturing, semiconductor, financial services, and energy sectors, with deep familiarity with the specific BCM challenges facing Taiwanese organisations in complex global supply chains. Our complimentary BCM diagnostic provides a structured, no-obligation starting point for organisations ready to build or strengthen their business continuity management capability.

積穗科研株式会社（Winners Consulting Services Co. Ltd.）は、台湾の事業継続管理（BCM）専門コンサルティング機関として、2018年に発表された工業プロセス最適化研究から重要な知見を提示します。煉油所の水素ネットワークにおけるリアルタイム最適化器（RTO）の検証研究は、不確実性の高い運用環境において、人間の判断に依存する協調的意思決定はシステマティックな最適化ツールに対して恒常的な性能差を示し、特に低負荷状況でその差が拡大するという事実を明らかにしました。この発見は、ISO 22301に基づく事業継続計画（BCP）における意思決定支援ツールの検証体制構築に直接的な示唆を持ちます。