How to Prepare a Decision-Ready Petroleum RFQ

How to Prepare a Decision-Ready Petroleum RFQ

ANFAXIS Energy Intelligence · Edition dated 23 September 2026

Executive takeaway

What Moroccan context makes this discipline relevant?

Why do fuel RFQs often create false price differences?

Because bidders may not price the same boundary. One may include final transport and waiting time; another may quote ex-terminal; a third may use an indexed formula with a different fixing lag. If procurement compares only the headline price, it ranks different commercial and logistical structures as if they were equivalent.

The RFQ’s first job is therefore to lock the comparison perimeter. Define delivery point, risk transfer, inclusions and exclusions, price formula, units and period. Require deviations in an exceptions matrix. The lowest-cost offer can only be identified after normalization.

Which product and volume data should the buyer provide?

Start with exact product, specification, unit, estimated volumes and frequency. Separate annual volume, monthly profile, daily peak and tolerance. An annual total without a profile can cause suppliers to mis-size storage, transport or credit. Disclose known seasonality and events that materially change demand.

Avoid false precision: a forecast remains an assumption. Provide relevant history, explain structural breaks and require bidders to state minimums, tolerances, nomination rules and rescheduling conditions. A good RFQ makes uncertainty visible so that it can be priced, allocated or managed.

How should delivery, Incoterm and delivered-cost boundary be defined?

Identify country, port or terminal where relevant, final site, delivery window, transport responsibilities and transfer point. The Incoterm does not replace the operational description: receiving constraints, opening hours, unloading means, documentation and waiting rules still need to be specified.

For decision purposes, request enough decomposition to rebuild delivered cost: product, differential, FX where relevant, freight/insurance, port/terminal, storage, inland transport, exception charges and payment terms. The objective is not necessarily to discover every supplier internal cost; it is to prevent different service scopes from appearing comparable.

What should the RFQ ask about storage, logistics and continuity?

State whether the site has storage, how much inventory is usable, what receiving constraints apply and whether upstream capacity is part of the requirement. Then ask for the normal logistics model, delivery throughput, mobilized resources, slots and disruption plan. Capacity without a pathway to the site is not evidence of continuity.

Define service levels that matter: complete/on-time delivery, replacement lead time, incident notification and reserved inventory or capacity only where contractually real. Obligations should be realistic and auditable. A penalty is not a substitute for a contingency pathway and should be assessed separately with competent advisers.

How can quality and traceability be included without bloating the RFQ?

Reference the approved specification, required certificates, transfer-point controls, sampling or inspection mechanism and non-conformance procedure. Article 07 covers the full evidence chain. The RFQ need not reproduce a quality manual, but it should allocate responsibilities and identify documents that condition receipt or payment.

Also ask how quantity variance, contamination, commingling and batch traceability are handled. A requirement with no evidence, threshold or owner will not help during a dispute. Applicable methods and standards should be validated by competent functions before tender release.

Which commercial terms can change the ranking?

Payment days, prepayments, guarantees, currencies, indexation, fixing windows, volume minimums and take-or-pay obligations can change economics. A unit discount can be erased by more cash tied up or insufficient flexibility. Article 15 deepens this issue; the RFQ should already make the terms comparable.

Require offer validity, market assumptions, adjustment mechanics and change governance. When a formula depends on an external index, specify source, window and timing. The buyer should be able to reproduce or audit the mechanics even without knowing the supplier’s margin.

How should an evaluation model avoid becoming a false score?

Separate minimum compliance from comparative criteria. A blocking regulatory, HSE or technical requirement should not be offset by a better price. For comparative criteria, define evidence and normalization rules in advance. Avoid subjective ratings without anchors: “supplier quality 8/10” is less useful than a traceable list of evidence and deviations.

The committee can then compare normalized delivered cost, continuity, quality, flexibility, payment terms and execution capacity. Document assumptions and exceptions. The decision record should explain why the selected offer is preferred, which risks remain and which conditions must close before contract signature.

What process should connect RFQ to award?

Before issue, have operations, quality/HSE, finance and legal review the package. During the tender, manage questions traceably and share material clarifications according to the chosen process rules. At bid receipt, check completeness first, then normalize offers to the same volume, timing and service perimeter.

After comparison, conduct clarifications without silently rewriting the requirement. Material changes should be documented. Before signature, reconcile the final offer with the contract: the value of a disciplined RFQ disappears if final clauses change responsibilities, inclusions or service levels without updating the evaluation.

How should this analysis be turned into a governed decision?

For RFQ and award, create one decision record shared by procurement, operations, finance, quality and legal. Class every material input as verified evidence, approved assumption, missing data or item requiring validation. This prevents an old assumption from becoming a fact simply because it appears in several presentations. Keep source, date, owner, version and next review date with each material input.

Separate blocking criteria from comparative criteria. An unmet regulatory, HSE, technical or compliance requirement should not be offset by a stronger economic result. Comparative criteria should use one boundary, coherent units and common assumptions. Where judgment is unavoidable, document the reasoning and the relevant counter-argument instead of hiding the trade-off inside an average score.

Define decision authority explicitly: who may request a study, commit spend, accept residual risk, approve a deviation and sign the contract or project decision. Escalation should be tied to observable triggers. Governance that is too vague slows response during disruption; governance that is too permissive can transfer risk into HSE, quality or finance.

Maintain a decision log. At each gate record the decision, evidence used, assumptions, conditions, accountable owner and revalidation date. When circumstances change, update the decision rather than silently editing the model. This traceability improves execution, auditability and learning across sites or transactions.

What should be completed in the first 90 days?

The 90-day rhythm is a governance proposal, not a regulatory requirement. An incident, market constraint or project schedule may require a shorter cycle; a major investment can require longer. Preserve the principle: close critical uncertainties progressively before increasing capital or risk commitment.

Then measure governance effectiveness: decisions supported by complete evidence, open gaps, closure time, expired assumptions, and incidents or changes that triggered revalidation. These indicators do not prove the final economic outcome, but they show whether the organization is actively managing the conditions that make the decision defensible.

Which failure modes should be tested before approval?

Run a pre-mortem: assume the decision has failed twelve months later and ask which assumptions proved wrong. Group potential causes into four families: incorrect or stale data, underestimated physical constraint, unmet commercial/regulatory condition, and governance unable to act in time. The exercise does not predict failure; it exposes dependencies that the central case can hide.

For each cause, identify an observable early-warning signal. A slipping lead time, unconfirmed capacity, utilization below plan, regulatory change, poor data quality or a counterparty failing to close a condition is more actionable than a single aggregate risk score. Link the signal to an action, owner and deadline.

Then test at least a lower-demand case, a higher-cost case and a delayed-schedule case. Where continuity or safety matters, add loss of a critical link. Where regulation matters, add a later permit or connection. Do not average scenarios that represent distinct blocking conditions; management needs to see which condition breaks the decision and why.

The final decision should state residual risk accepted and the conditions that trigger re-approval. This prevents a project or contract from continuing by inertia after its economic, technical or regulatory logic has changed. A good decision pack does not eliminate uncertainty; it shows where uncertainty sits, who monitors it and which event would change the decision.

Checklist — 12 fields to lock before issue

Approved product and specification

Unit, annual volume and monthly profile

Peak, tolerance and delivery frequency

Destination, transfer point and delivery window

Incoterm and additional responsibilities

Existing storage and required capacity

Access and unloading constraints

Price formula, index, window and currency

Quality, sampling, quantity and claims

SLA, backup and replacement lead time

Payment, guarantees and volume flexibility

Exceptions matrix and clarification governance

Frequently asked questions

Should the RFQ disclose a budget?

Not always. A budget may help qualify some solutions but can also anchor responses. For standard supply, requirement and formula comparability matter more. For a package including infrastructure, storage or additional services, an approved budget band can be useful.

Can the buyer request a fixed price?

Yes if the supplier accepts and the period is realistic, but clarify what is actually fixed: product, differential, freight, FX or only selected components. An apparently fixed price may carry exclusions or a very short validity. Ask for revision rules and validity.

How many suppliers should be invited?

There is no universal number. It depends on procurement strategy, market structure, internal rules and qualification requirements. The objective is credible competition among capable bidders, not a high response count with poor comparability.

Can an RFQ serve as the contract?

It can feed the contract but should not be assumed to replace the final legal agreement. Responsibilities, warranties, remedies, force majeure, limitations, governing law and other clauses require formal drafting and competent review.

Should supplier licences be requested?

Request evidence relevant to the scope actually performed and verify it through appropriate authorities or sources. Do not assume that a commercial document or stated capacity proves the authorization required for the transaction.

How do buyers reduce incomplete bids?

Use a mandatory compliance table and exceptions matrix, then reject or clarify missing responses before economic scoring. The evaluation should distinguish missing information, explicit deviation and non-compliance.

From insight to decision

Sources and methodology

Research cutoff is 23 September 2026. Market data come from the Moroccan Ministry; the RFQ structure is ANFAXIS editorial analysis aligned with the website workflow. It is not legal advice, a live tender or a commercial commitment.

[1] Morocco Ministry of Energy Transition — Fuels, 2026 key indicators / 2025 consumption: mem.gov.ma ↗

[2] Ministry — Procedures and authorisations: mem.gov.ma ↗