Reserves classification — a systematic guide to PRMS

Two engineers look at the same field and report different numbers — not because either is wrong, but because they are answering different questions. How much petroleum is there? How much can we get out? How much can we sell, profitably, starting now? Reserves classification is the discipline that keeps those questions separate and makes the answers comparable across companies, countries, and decades. This is a complete walk through that framework, built on the petroleum industry’s global standard: the SPE Petroleum Resources Management System.

Why classification exists

A reserves number is never a measurement; it is an estimate of the future made under uncertainty and tied to economics that change. Left undisciplined, the word “reserves” becomes meaningless — one company’s optimistic guess is another’s booked asset, and an investor cannot compare them. Worse, reserves underpin real money: company valuations, bank loan collateral (reserve-based lending), national resource accounting, and securities disclosures. A shared, rigorous classification system exists so that the same volume means the same thing to a reservoir engineer, an auditor, a lender, and a regulator.

The international standard is the Petroleum Resources Management System (PRMS). First issued in 2007 and substantially revised in June 2018 (with errata in 2018 and 2022), PRMS is sponsored by the Society of Petroleum Engineers together with the World Petroleum Council, the American Association of Petroleum Geologists, the Society of Petroleum Evaluation Engineers, the Society of Exploration Geophysicists, the European Association of Geoscientists and Engineers, and the Society of Petrophysicists and Well Log Analysts. It is used for portfolio management and for national regulatory disclosure in many jurisdictions, and it supplies the petroleum-specific specifications under the United Nations Framework Classification for Resources (UNFC).

The mental model: in place, recoverable, commercial

Everything starts with what is physically present and narrows from there. The total volume of hydrocarbons originally in the rock is the Petroleum Initially In Place (PIIP). Only a fraction of that can ever be brought to surface with any technology — the rest is unrecoverable. Of the recoverable portion, only some is commercial under today’s prices, costs, contracts, and regulations. Reserves live at the innermost ring: discovered, recoverable, and commercial, with what is left to produce from a committed project.

The two-axis framework

The heart of PRMS is a single picture with two independent axes. The vertical axis is the chance of commerciality — how likely it is that the volume will actually be developed and sold. The horizontal axis is the range of uncertainty in the recoverable quantity itself — the spread from a conservative low estimate to an optimistic high one. Every volume on earth occupies one cell in this grid. Move up the vertical axis and a project matures from a geological idea toward producing cash flow; move left to right and you go from what you are confident about to what is merely possible.

The three classes

Reserves are quantities anticipated to be commercially recoverable from a given date forward, by application of development projects to known accumulations. Four conditions must all hold: the accumulation is discovered, the volume is recoverable with current technology, the project is commercial, and the quantity is what remains to be produced. Crucially, in PRMS reserves attach to a project — a defined development with a committed path to market — not merely to a volume of rock.

Contingent Resources are discovered and potentially recoverable, but not yet commercial. The petroleum is there and you could in principle produce it, but something blocks the commercial decision: no market or evacuation route, an immature technology, a pending government approval, or economics that do not yet clear the hurdle. The moment those contingencies are resolved and a development is committed, Contingent Resources reclassify upward into Reserves.

Prospective Resources are estimated volumes associated with undiscovered accumulations — the exploration upside. They carry an additional risk the other two classes do not: a chance the accumulation does not exist at all (geological chance of success). A successful exploration well moves a prospect’s volumes from Prospective into Contingent Resources, pending the commercial decision.

Reserves by uncertainty: Proved, Probable, Possible

Within Reserves, the horizontal axis splits the estimate into three incremental categories of decreasing confidence. Proved (P1) reserves are those that analysis suggests have a high degree of confidence of being recovered. Probable (P2) are additional reserves less certain than Proved but more likely than not. Possible (P3) are additional reserves less likely than Probable. These are usually quoted as cumulative scenarios:

The categories are defined two ways, which should give consistent answers. Under the deterministic method, “reasonable certainty” for Proved means a high confidence the actual recovery will equal or exceed the estimate. Under the probabilistic method, the bar is explicit: there should be at least a 90% probability that quantities recovered equal or exceed the 1P estimate, at least 50% for 2P, and at least 10% for 3P. In the petroleum convention these are exceedance probabilities, so 1P ≈ P90, 2P ≈ P50, and 3P ≈ P10 — the low, median, and high of the recoverable distribution.

Reserves by status: developed and undeveloped

The same Reserves are also cut a second, independent way — by how ready they are to flow. Developed Producing reserves come from completion intervals open and flowing at the evaluation date. Developed Non-Producing reserves sit behind pipe or in shut-in zones that need only minor work (a recompletion, opening a zone) to produce. Undeveloped reserves require significant new investment — new wells, major recompletions, or new facilities — before they can be produced. The split matters because developed reserves are closer to cash and carry less execution risk than undeveloped ones, even when both are equally “Proved.”

Project maturity sub-classes

Because PRMS attaches volumes to projects, each class is further divided by how mature the project decision is. This is the vertical axis made concrete, and it is where reserves work meets the real calendar of approvals and budgets.

For Reserves, the sub-classes are On Production (the project is producing), Approved for Development (capital is committed and execution has begun), and Justified for Development (the project is economically and technically sound and a final investment decision is expected imminently). For Contingent Resources, the ladder is Development Pending, Development On Hold or Unclarified, and Development Not Viable. For Prospective Resources, the units are Prospect (a drill-ready target), Lead (a less-defined target needing more work), and Play (a prospective trend or fairway).

The commerciality test in detail

The line between Contingent Resources and Reserves is the most consequential boundary in the system, because crossing it turns a resource into a bookable asset. PRMS sets several conditions for a project to support Reserves. There must be evidence of a firm intention to develop within a reasonable timeframe — the guideline is generally about five years, though longer horizons can be justified for large or phased projects. The development must be economic under a defined forecast of prices and costs. All necessary approvals, markets, and evacuation infrastructure must be in place or reasonably expected. And the volumes must be measured to a defined reference point — usually the custody-transfer or sales point — so that everyone counts the same molecules at the same place.

Two refinements matter for reporting. Gross versus net: gross reserves are the total for the project, while net (entitlement) reserves are the share that accrues to a particular company under its fiscal or contractual terms — a distinction that can move a number dramatically under a production-sharing contract. And reserves are always quoted with their economic cut-off: production stops being counted once the project stops covering its operating costs, even if petroleum physically remains.

Deterministic and probabilistic estimation

PRMS permits two estimation philosophies, and good practice often runs both as a cross-check. The deterministic approach builds discrete scenarios — a single Proved case, a single 2P case — from specific input values chosen to reflect the required confidence, either by the incremental method (adding categories of increasing risk) or the scenario method (constructing full low/best/high cases). The probabilistic approach assigns distributions to the uncertain inputs — area, thickness, porosity, saturation, recovery factor — and propagates them, typically by Monte Carlo simulation, into a full distribution of recoverable volume from which P90, P50, and P10 are read directly. Neither is inherently superior; the probabilistic view simply carries the uncertainty information the deterministic scenarios summarize.

Aggregation: the trap of adding up

A subtle but critical issue arises when combining reserves across wells, reservoirs, or fields. The instinct is to add the categories — sum all the Proved volumes to get field Proved, sum all the 3P volumes to get field 3P. This arithmetic summation is systematically biased. Adding many independent conservative Proved estimates produces a total that is more conservative than the true field P90, because it is statistically unlikely that every single entity simultaneously hits its low case. The reverse happens at 3P: summing many independent high cases overstates the field P10, since not everything hits its high case at once.

Probabilistic aggregation — combining the distributions rather than the point estimates — captures this portfolio effect correctly. As a rule of thumb, for independent entities the arithmetic sum of Proved is lower than the probabilistically aggregated Proved, and the arithmetic sum of 3P is higher than the probabilistic 3P; the 2P best estimate is the least affected. The practical guidance: aggregate at the lowest sensible level, be explicit about whether entities are independent or correlated, and never quote a field-level P90 that is simply the sum of well-level P90s without noting the conservatism.

PRMS, SEC, and UNFC — how the systems relate

PRMS is the broad management standard, but it is not the only system a reserves number may have to satisfy. In the United States, companies filing with the Securities and Exchange Commission (SEC) must report under SEC rules, which are narrower and stricter for public disclosure. The SEC’s modernized rules (effective 2009) primarily recognize Proved reserves for headline disclosure, mandate a specific price basis — the unweighted twelve-month average of first-of-month prices — rather than a forward forecast, permit reliable technology and probabilistic methods, and impose a five-year rule on Proved Undeveloped locations. The result is that the same field can carry a larger 2P number under PRMS for internal and lender purposes and a smaller Proved number under SEC for the annual report.

At the other end, the United Nations Framework Classification (UNFC) is a high-level, system-neutral scheme designed to harmonize all energy and mineral resources across countries. PRMS serves as the petroleum-specific specification that maps into UNFC, so a volume classified under PRMS can be expressed in UNFC terms for national resource accounting. The takeaway for any evaluator: always state which system a number is reported under, because “reserves” alone does not pin down the meaning.

Common pitfalls

A handful of errors recur in reserves practice. Treating reserves as a property of rock rather than of a committed project leads to booking volumes that have no path to market. Confusing the statistical and petroleum conventions for P10/P90 flips the optimistic and conservative ends. Aggregating by naive arithmetic summation and quoting the result as a field P90 or P10 ignores the portfolio effect. Mixing reporting bases — comparing a PRMS 2P figure against an SEC Proved figure as if they were the same — produces meaningless conclusions. And forgetting the reference point or the gross-versus-net distinction can misstate entitlement by a wide margin. Every one of these is avoided by the same discipline: state the class, the category, the method, the reporting system, and the reference point, every time.

Closing

Reserves classification looks bureaucratic until you see what it protects: the ability to compare one company’s future production against another’s honestly, and to attach money to those numbers responsibly. PRMS does this with two ideas held together — how likely a volume is to be developed, and how uncertain the recovery is — and a vocabulary precise enough that an engineer, an auditor, and a banker mean the same thing by the same word. Learn the matrix in Figure 2, internalize that reserves belong to projects, respect the aggregation math, and name your reporting system, and you can speak the language of reserves with confidence.

References
SPE / WPC / AAPG / SPEE / SEG / EAGE / SPWLA (2018). Petroleum Resources Management System (PRMS), Revised June 2018. Society of Petroleum Engineers.
SPE et al. (2011). Guidelines for the Application of the Petroleum Resources Management System.
SPE Oil & Gas Reserves Committee (2018 & 2022). PRMS Errata.
U.S. Securities and Exchange Commission (2009). Modernization of Oil and Gas Reporting; Final Rule. 17 CFR Parts 210, 211, 229, 249.
UNECE (2019). United Nations Framework Classification for Resources (UNFC) — Update 2019.
Ahmed, T. (2010). Reservoir Engineering Handbook, 4th ed. Gulf Professional Publishing — reserves definitions and classification.