EL7.AI
Strategy BuilderCOT DataAdvanced NewsResearchNewEarnings CalendarEconomic Calendar
We use cookiesPrivacy Policy
EL7.AIEL7.AI

AI Financial Intelligence
Professional analysis of central bank decisions

© 2026 EL7.AI. All rights reserved.

Markets

  • News
  • Forex
  • Stocks
  • Crypto
  • Gold
  • Commodities
  • Indices
  • ETFs

Analysis

  • Fed
  • ECB
  • COT
  • Economic Calendar

Learn

  • Service Guide
  • Oil

Company

  • About
  • Contact
  • Data Methodology
  • AI Disclosure
  • Pricing
  • Enterprise
  • Terms
  • Privacy
  • Security
  • WhatsApp
Status data is currently unavailable

The information provided on EL7.AI is for educational and informational purposes only and does not constitute financial advice.

EIA Weekly Refinery Utilization Rates WoW

United StatesEnergyWeekly%Low
SummaryAbout This Indicator4Historical Data121
Latest release · October 15, 2026 · 16:00 UTC
0.80%
No forecast was recorded for this release
Last five prints
-1.00-2.80-1.500.200.80

The reading has risen in each of the last 3 releases.

Forecast
—
Previous
0.20Actual is higher by 0.6

About This Indicator

4 questions

The Weekly U.S. Percent Utilization of Refinery Operable Capacity measures weekly refinery throughput divided by operable capacity. Throughput means crude oil, unfinished oils and plant liquids run through distillation units, while operable capacity means ready distillation capacity. It is one series inside the Weekly Petroleum Status Report from EIA, the statistical and analytical agency within the U.S. Department of Energy. It covers U.S. operable refineries, not the count of plants running, demand, gasoline output or storage use.

Weekly data usually cover Friday 7:01 a.m. through the next Friday 7:00 a.m., with Monday 5:00 p.m. as the submission deadline. Releases usually fall Wednesday at 10:30 a.m. Eastern Time for tables and highlights, then 1:00 p.m. for the full report. Values describe the week ending the prior Friday morning, with holiday weeks delayed by one day. Each release pairs the rate with input and output tables by PADD district and rarely revises past weeks unless a fix changes supply understanding.

Moves come from throughput changes, capacity updates and related oil flows.

  • Gross inputs: crude oil, unfinished oils and plant liquids run through distillation units in the week.
  • Operable capacity: latest monthly ready distillation capacity, so it shifts only when monthly data shift.
  • Crude inputs and output: weekly input volumes plus gasoline and distillate production explain throughput shifts.
  • Regional splits: national and PADD district results show where throughput changed.

For the week ending August 10, 2018, inputs averaged 18.0 million barrels per day, up 383,000, while utilization reached 98.1%. Commercial crude stocks still rose 6.8 million barrels to 414.2 million as imports jumped 1,083,000 barrels per day to 9.0 million. Gasoline stocks fell 0.7 million barrels while distillate stocks rose 3.6 million, as gasoline output averaged 10.2 million barrels per day and distillate 5.3 million.

For the week ending August 17, 2018, inputs slipped 89,000 barrels to 17.9 million barrels per day while utilization held at 98.1%. Crude stocks then drew 5.8 million barrels to 408.4 million.

  1. Weekly sample: EIA collects Form EIA-800 reports from a cut-off sample of the largest firms covering about 90% of volumes. Missing reports use exponential smoothing, or weighted past values, of each firm's prior weeks.
  2. Ratio expansion: EIA scales the weekly sample to all firms using lagged monthly census totals. The weekly total equals the weekly sample sum times the ratio of the all-firm monthly total to the sample monthly total. EIA may adjust the ratio for unusual events missing from lagged monthly data.
  3. Weekly inputs: This method directly estimates total gross inputs to refineries and production.
  4. Operable capacity: EIA uses the latest reported monthly operable capacity as the base.
  5. Utilization ratio: Utilization equals 100 times gross inputs divided by operable capacity with unrounded numbers. For example, 16,500 divided by 18,000 times 100 equals 91.7%, rising to 93.3% if inputs reach 16,800, a rise of 1.6 percentage points.
  6. Review: The review team checks totals before Wednesday release and requires resubmission for errors above 5%. EIA publishes revisions only for substantive supply effects. Lagged monthly capacity overstates capacity and understates utilization until shutdowns enter monthly data.

Key point

98% means throughput over capacity, not 98% of plants running, a demand gauge or storage use. It tracks wholesale movement, not final consumption.

Historical Data

View data as table
DateActualForecastSurprise
Oct 15, 20260.80——
Oct 7, 20260.20——
Sep 30, 2026-1.50——
Sep 23, 2026-2.80——
Sep 16, 2026-1.00——
Sep 10, 2026-0.20——
Sep 2, 20260.60——
Aug 26, 20260.20——
Aug 19, 20261.00——
Aug 12, 2026-0.30——
Aug 5, 2026-0.70——
Jul 29, 20261.10——
Jul 22, 2026-0.10——
Jul 15, 20260.40——
Jul 8, 2026-0.80——
Jul 1, 20260.50——
Jun 24, 2026-0.60——
Jun 17, 20261.40——
Jun 10, 20260.60——
Jun 3, 20260.20——
May 28, 20262.90——
May 20, 2026-0.10——
May 13, 20261.60——
May 6, 20260.50——
Apr 29, 20260.50——
Apr 22, 2026-0.50——
Apr 15, 2026-2.40——
Apr 8, 2026-0.10——
Apr 1, 2026-0.80——
Mar 25, 20261.50——
Mar 18, 20260.60——
Mar 11, 20261.60——
Mar 4, 20260.60——
Feb 25, 2026-2.40——
Feb 19, 20261.60——
Feb 11, 2026-1.10——
Feb 4, 2026-0.40——
Jan 28, 2026-2.40——
Jan 22, 2026-2.00——
Jan 14, 20260.60——
Jan 7, 20260.00——
Dec 31, 20250.10——
Dec 17, 20250.30——
Dec 10, 20250.40——
Dec 3, 20251.80——
Nov 26, 20252.30——
Nov 19, 20250.60——
Nov 13, 20253.40——
Nov 5, 2025-0.60——
Oct 29, 2025-2.00——

Indicator questions

Source · U.S. Energy Information AdministrationView Source
Next release
Wednesday, October 21 · 14:30
4 days
October 15October 21

Statistics

121 readings
  • Mean-0.03
  • Std Deviation1.26
  • TrendRising
  • Year-over-year change+1.10
Latest reading against its usual range (mean ± one standard deviation)
0.80-1.29-0.031.23

The latest reading is within its usual range, less than one standard deviation from the mean.