Propane Draws 2.7M; Brutal Cold Grips Much of America

The EIA announced a propane draw of 2.7M/bbls for the week ending 12/22/17, and I am not buying it. Why not? Export totals were listed at just 920,000/bpd, which was a drop off from last week’s 1.202M/bpd.

We subscribe to a waterborne export service that while not always accurate, rarely misses the way it ‘missed’ this week. Two weeks ago, their estimate was within 25,000/bbls of the total number, which is nearly a direct hit. Last week’s estimates were also in the ball park of usability as a predictive tool. The report suggested we’d see 1.5M/bpd of exports this week, which would have been an all time record and an estimate I was skeptical of because I don’t know that we can theoretically export that much propane in one week.

However, I was thinking we’d see no worse than last week’s 1.202M/bpd and possibly as high as 1.35M/bpd. If we had matched last week’s totals, our draw would have been 4.7M/bbls, which was more in line with our internal thinking. Instead, exports were listed at 920,000/bpd, a decrease of nearly 300,000/bpd week over week.

I am not mad because my ideas were rebuffed…I am irritated when the data makes no sense and appears to be at odds with itself…it just smells like someone failed at some mathematical computation here. But in the end, none of that matters; the report is what it is, and if there is a correction to make in the data down the road, they will make it in the monthly totals and not the week to week data reports which have a much greater impact on market direction.

As such, this 2.7M/bbl propane draw will likely have a dampening effect on prices, pushing them a little lower today than where they closed yesterday.

One year ago this week, we saw a draw of 5.6M/bbls, so our Year over Year deficit is now down to 18.3M/bbls. That’s still a big gap, and we drew down 7.2M/bbls last winter between 12/23/16 through 1/6/17. I think we will exceed that total this year.

I am not rooting for chaos, although I enjoy challenges. I just like data outputs to make sense with the data inputs…and today’s report didn’t do that.

WEATHER: Welcome to one of the coldest 15 day periods since at least 1950 in the Midwest and in the East. We are hearing of red-line loading at some terminals along the Dixie Pipeline as well.

According to MDA, nine of the next 10 days are among the ten coldest from a Gas Weighted Heating Degree Day standpoint for their respective dates, in some Midwestern and Eastern areas, in the last 67 years!

Take a look at this, again from MDA: “The current 10 Day forecast totals 405.88 GWHDDs, which is colder versus the previous period record of 390.88 GWHDDs from 1967-68.”

And then this: “The period from December 27 to January 5 at 409.9 GWHDDs is forecast to be the coldest 10 Day stretch dating back to January 1997 (reference of past few winters: 2016-17 season peaked at 326.4; 2015-16 season peaked at 330.8; 2014-15 season peaked at 380.7; 2013-14 season peaked at 384.8). The December 25 to January 8 period at 586.7 GWHDDs is forecast as the coldest 15 Day stretch since January 1994 (reference of past few winters: 2016-17 peak at 475.8; 2015-16 peak at 490.5; 2014-15 peak at 563.6; 2013-14 peak at 551.0).”

Yes, colder than the same stretch in 2013-2014. Also, don’t get bogged down in what those numbers actually mean, as Gas Weighted Heating Degree days differ from your standard understanding of Heating Degree Days. Instead, focus more on the percentage differences. The projected 586.7 GWHDD between December 25th through January 8th is over six percent colder than the frigid 2013-2014 time frame, and colder than the same time frame in 2014-2015 that was actually colder than the 2013-2014 spell.

The coldest air of the winter of 2013-2014 came in mid-January, but what we are experiencing and will experience over the next 10 or so days is historically cold.

This level of cold is almost too cold…the trucking industry is beginning to buckle under this level of demand. The E-Log system is causing significant problems for an industry that has voluntarily contracted over the past few years, due to warmer winters. Some companies have cut staff as well as rolling stock, and I know of a few multi-state trucking companies that are scheduling three and four days out.

There may be a period of transient warmth around January 8th or so, but that ‘warm up’ or at least it’s severity is being cast in doubt more and more with each passing day.

Two days ago, the European model was projecting a trough off the coast of Alaska. The downstream effects from something like that typically means milder air east of the Mississippi. Then yesterday’s same computer model projected a ridge off the coast of Alaska, which leads to colder air downstream, east of the Mississippi. Or to put it simply, it projected the exact opposite from one day to the next.

It still looks like a milder period could be coming around January 8th through the 18th, but the models are struggling with consistency and they are showing a Southeastern Ridge, which they showed at times for December, and that ridge never materialized and the cold air won out. Some meteorologists, like Michael Clark from, is questioning whether that ridge will be there in 20 days. The models are already beginning to push the ridge farther out into the Atlantic ocean.

If the latter happens, the ‘warm up’ will be limited in both its intensity compared to normal temps as well as its duration.

Here is an example of why Clark is dubious of the Southeastern Ridge and therefore the models. The first picture is what the European Model projected for Christmas Day, nine days in advance, on 12/16:

Note the warmth in the Southeast, which was a prediction of a Southeastern Ridge, and all of that mild air, and an elongated cold swath making its way down into the Southwestern corner of the United States. Now, here is what actually happened on Christmas Day:

The Southeastern Ridge was pushed out into the Atlantic…the Southwest did not have a low pressure trough and thus colder air, it also had a ridge. Even in their errors the models can offer clues. In some years their biases, if consistently wrong in the same way, tells you they are struggling to take into account other atmospheric factors. You can then anticipate their errors and have an idea of what is going to happen. We have seen this same error at least twice in the last 30 days, and present modeling is beginning to suggest we could see it again.

We all love cold weather in this industry, but sometimes, we can have too much of a good thing, given the logistical constraints our industry has to deal with on the trucking side and sometimes on the pipeline side.

You have have an ocean of water at your disposal, but if you only have a garden hose as your delivery mechanism, it takes a long time to move the water. That has been our industry’s biggest problem since I began wholesaling in 1996 and it remains our biggest challenge.

Brutal cold is here and will be with us for a time…we may not get much time to catch our collective breath in a few weeks. Fun times, but challenging times.

LAST THING: I just subscribed to a service that allows me to go in and look at the weather models of my choosing, and download GIF’s…I hate to admit how excited I am about this, but geek is gonna geek. Take a look at these temperature anomalies (departure from normal) through January 7th!

I like it…both the cold and this new weather tool I have at my disposal!

Jon Miller
Marketing Representative for NGL Supply Wholesale in Tulsa Oklahoma. Follow me on twitter @PropaneBuzz

Join Our Newsletter

  • Tweets

Copyright © 2018 Propane Buzz | All Right Reserved