Natural gas storage: separate inventory, capacity and daily deliverability

Three questions hidden inside a storage headline

A natural gas inventory number tells you how much gas is stored at a particular time. It does not, by itself, tell you how much the facilities could hold or how quickly they could supply customers. Those are three different research questions: inventory, capacity, and deliverability. Confusing them can make a reassuring national total look like proof that every local market has enough immediately available supply.

This is an educational framework, not a report of a new weekly storage reading. The examples are hypothetical. The purpose is to make a storage claim testable before using it in research on an energy producer, utility, or infrastructure business. Start with the measurement, then identify the operational and commercial evidence needed to connect it to a company.

Separate working gas from the operating cushion

EIA distinguishes base gas, intended to remain in a reservoir to support pressure and deliverability, from working gas above that level. Working gas is available to the market. Working gas capacity is total storage capacity less base gas. These definitions come from EIA's storage explainer; its historical industry counts should not be mistaken for current figures.

Suppose a fictional facility has total capacity of 150 units, base gas of 50, and working inventory of 70. Its working capacity is 100 units, so its working inventory occupies 70 percent of working capacity. Total gas physically present is 120 units, which is 80 percent of total capacity. Both percentages are arithmetically correct; they answer different questions.

Before writing that a facility is 80 percent full, specify which fraction you mean. Otherwise, a reader may assume that all the gas counted in the numerator is routinely available for sale or that the denominator measures usable working capacity. In this example, neither assumption follows from the 80 percent figure. Keep the operating cushion visible in the calculation.

An observed peak is not the same as a design limit

EIA's capacity definitions distinguish working gas design capacity from demonstrated peak capacity. Design capacity reflects the reservoir, equipment, and operating procedures. Demonstrated peak capacity adds each field's highest reported working inventory over the stated five-year window, even when those individual peaks occurred at different times. The definitions page reviewed here covers data through November 2025 and was released in May 2026.

That timing detail changes how a researcher should read the total. It need not correspond to a single historical date on which all fields simultaneously held their maximum inventories. Nor does a historical peak establish how much a facility can deliver to a particular customer tomorrow.

For a company presentation, copy the exact capacity label into your notes. If management announces an expansion, ask whether the number describes additional working volume, additional withdrawal capability, or another measure. Do not compare it with an older figure until you have checked the definition, facility boundary, and effective date. A percentage increase without those details may conceal a change in what is being measured.

Plenty of volume can still leave a daily bottleneck

EIA describes deliverability as the rate at which gas can be withdrawn, commonly measured per day. It depends on operating conditions including pressure, gas remaining in the reservoir, and compression capability. The storage explainer notes that deliverability generally decreases as working gas is drawn down. Treat a constant-rate calculation as a simplified scenario, not an operating guarantee.

Imagine a separate fictional facility containing 60 units of working gas. A customer needs three units per day, but the assumed withdrawal limit is two units per day. Dividing inventory by demand gives twenty days of volume coverage. Yet the facility cannot satisfy even the first day's entire requirement under that withdrawal constraint. There is a one-unit daily gap, despite enough inventory on paper.

At a constant two-unit daily withdrawal, emptying sixty units would take thirty days. That is merely the arithmetic of the assumed rate. Actual capability could change as inventory falls. The example shows why volume coverage and daily supply capability should occupy separate rows in a research model. It does not forecast any real facility's performance.

Use the historical average for the question it answers

EIA's weekly methodology describes estimated working inventories for the Lower 48 states and regions. It also explains how interpolated daily historical estimates support comparisons with the same calendar dates in earlier years. A five-year average is a historical benchmark; it is not the physical capacity of the storage system.

Consider hypothetical working inventory of 3,200 units, a comparable historical average of 3,000, and separately established working capacity of 4,000. Inventory is 200 units, or about 6.7 percent, above the historical average. It is also 80 percent of the assumed capacity. Saying above average and saying nearly full are different claims, requiring different denominators.

Our analytical recommendation is to preserve the official comparison supplied with the release rather than constructing an accidental comparison between mismatched weeks. Also check the geographic boundary. A broad inventory total can be useful context while still leaving a specific facility's availability unresolved. Move from the aggregate to the relevant operating area before making a company claim.

Build the bridge to a business without inventing an outcome

For an infrastructure operator, investigate whether contracts pay for reserved capacity, actual activity, or a combination. For a gas producer, investigate the relevant delivery location, transport arrangements, and disclosed price exposure. For a utility, investigate access to storage and other supplies during the period being discussed. These are proposed research questions, not assertions about any named company's agreements.

A national inventory observation does not answer those questions. The evidence bridge should connect the storage measure to a specific operational exposure and then to the applicable commercial terms. Without those steps, predicting revenue or margins from a storage headline skips the part that determines who bears the risk.

Record a countercase as well. Additional inventory could be accessible in the wrong place or on terms unavailable to the business you are studying. Conversely, a business might have alternative supply arrangements that reduce its dependence on the facility under discussion. These possibilities are reasons to check disclosures, not reasons to assume that either outcome has occurred.

A practical worksheet for the next storage report

Write down the inventory date, publication date, geography, units, and whether the figure refers to working or total gas. Put the weekly change and historical comparison alongside it, but keep capacity and deliverability in separate fields with their own sources and dates. Leave an unknown field empty rather than deriving it from an unrelated measure.

Then state one testable interpretation and the next evidence required. For example, an apparent supply buffer may justify checking local withdrawal capability; it does not establish that capability. A claim about company earnings needs an additional review of the business's actual exposure. This keeps the worksheet useful when a later report changes the inventory number.

The cover uses stylized tanks and connecting pipes as a conceptual image. It is not a map of actual facilities or a technical diagram of underground storage. The distinction mirrors the article's main discipline: label what each piece of information represents before using it as evidence.