Thursday, August 22, 2019

Another US Crude Oil Production Record Set This Past Week -- August 22, 2019

The background to this note involves a lot of history about how the US government comes up with their statistics. If you've been following the blog for awhile you are probably aware of the "inside" story. If not, the story will become clear(er) over time.

From a reader:
The lower 48 set a production record this week, but I doubt that anyone mentioned it.

The output from wells in the lower 48 states rose by 100,000 barrels per day to a rounded record high of 12,000,000 barrels per day, while there was a 94,000 barrels per day decrease to 339,000 barrels per day in Alaska's oil production, and hence the final rounded national production total that everyone cites was unchanged at 12,300,000 barrels per day.
Most of this year, Alaska was adding 500k to the national total. I have no idea what happened in there this week, unless they had a hurricane we didn't hear about.
My hunch: polar bears are turning off the oil pumps to do their part to prevent global warming. 

This information will be repeated in greater detail and explanation this next Sunday, at "Focus On Fracking," linked at the sidebar at the right.

See EIA data at this link. And, yes, 12,300,000 million bopd for the week ending 8/16/2019. I believe those are "estimates."

This EIA data is apparently "actual" production data but such data lags a couple of months.

5 comments:

  1. Thank for clarity at the end. The commenter is citing weekly "data". This data is just an estimate, not a real survey. They actually take the STEO (Short Term Energy OUTLOOK) and just interpolate the forecasts. The STEO is based on the most recent EIA 914 (the real monthly data). That is generally 2-3 months ago. (The STEO is revised each month, but it just has the most recent 914 to work with.)

    Situation is very comparable to ND, where the Director's Cut is a month and a half old. If someone cited "current" ND production, it would have to be a modeled estimate (really a forecast).

    EIA even went to rounding off the weekly numbers to dissuade people from treating them as real. I wish they would get rid of them entirely. Their monthlies, like the NDIC Director's Cut are quite good. The weekly production is a joke though. (Weekly inventories are decent, but they are off of a real survey...are actual observations, not a model.)

    Another issue with the weeklies is they are NEVER updated. So if the weeklies are running low or high, they will catch up with a big jump. But they won't go back and fix old values. People routinely misinterpret this as a single week change, but it's not. It's "rebaselining".

    Similarly, if you show a time series of weeklies over time, it's not a good idea (since the old values are never fixed, you are just looking at the record of old predictions, not old data...and for the old times, you actually HAVE good data (the monthlies). What I prefer to see, but rarely do, is the monthlies in ONE COLOR...and then if you MUST show the last couple months, just use the STEO itself--in another color and with clear legend that it is a forecast. Note, this is basically the same as the weeklies, because all they do with the weeklies is just interpolate the monthly STEO values (now rounding them to make it even more obnoxious).

    The confusion with the weeklies is on of those recurring annoyances...that I shouldn't let bother me. But still. It's like people confusing BOE and BO. Grr.

    ReplyDelete
    Replies
    1. 1. Wow, you are a fast typist.

      2. I think it was you who first corrected me (months ago) that the weeklies were estimates and the monthlies were "actual."

      3. The failure of bureaucrats to update old information or to make it more clear that these are only estimates is very frustrating, and based on "job" data released this month, it sounds like the entire government has this problem. Estimates are way off. To get the final GDP, they have three readings (or thereabouts), as just one example.

      Delete
    2. They actually do clarify it fine. You just have to crawl down out of the cockpit, walk into the crew chief's area and look at the tech manual.* In this case, the FAQ. It's all in there. They've run webinars too.


      *Don't hate me, man...I roomed with an F-14 pilot in SD. I know the beast.

      Delete
    3. I think this is how Boeing got in trouble with the MAX 737.... they said it was in the manual also. LOL.

      Delete
  2. for the record, i pulled that data off the "U.S. Petroleum Balance Sheet", part of the Weekly Petroleum Status Report: https://www.eia.gov/petroleum/supply/weekly/ see "crude oil supply"
    that's also where i get the weekly "unaccounted for crude" number, line 13...as i've noted, that number, which represents a weekly margin of error, occasionally has week to week swings as great a 1 million bpd...

    so yes, we all know these numbers are estimates and often terribly incorrect, but these are the numbers that move oil prices every week, and hence almost everything else that we'd be interested in, including decisions to drill or complete a well...

    even worse, if i didn't write about these weekly numbers. i wouldn't have anything to blog about..

    ReplyDelete