Natural Gas Production, Use, Storage, & Price Analysis

(photo by sidewalk flying's)

On a post last week looking at crude oil, gasoline, fuel oil, and natural gas prices, I made the statement that natural gas did not correlate to crude oil like the others, there seemed to be something else going on with natural gas prices so I wanted to look further. EIA has a ton of great historical data on various energy sources.

Natural Gas Balance of Supply and Demand:

I will start by showing an interesting perspective of the supply and demand of U.S. natural gas over time since 2001:


What is immediately apparent is the strong seasonality of demand which makes perfect sense as much more natural gas is used in the winter months. "Supply" here is the sum of U.S. production plus imports minus exports. Recently the U.S. imports about 16% of the natural gas supply. Note that on an instantaneous basis there is not enough gas supply in the winter months to keep up with demand. This situation makes storage capability critical.

Natural Gas Storage Capacity Utilization

Natural gas reserves are typically stored underground in natural geologic reservoirs such as hollowed out salt domes, and depleted oil and gas fields. A fairly large amount (~65%) of the total storage volume is not usable. This base volume is needed for maintaining working pressure and for other process needs. The "working volume" is the amount really usable as recoverable storage. The plot below shows both total storage volume and working volume both as a % of total storage capacity over time since 2001.



Strong seasonality here which obviously follows from the seasonality in demand shown above. Storage hits a high in October and a low in March. Working storage typically gets above 80% in October and then falls to 40% or less in March. You can imaging what would happen should the working storage levels get near 0% with winter months remaining.

Natural Gas Price and Storage Volume Utilization

Below is a plot showing the price of natural gas and the working storage % utilization since 1992. There appears to be a relationship of inadequate stores in October leading to spikes in natural gas prices into the winter months. I drew a somewhat arbitrary line at 80% working storage and noted in blue filled circles Octobers which peaked below 80%. (click on any of the charts for a larger view)



Doing a rough eyeball analysis, I count 5 out of 6 occurrence of less than 80% working storage in October followed by a winter price spike in natural gas price (the p-value of this observation is 0.1 which is short of the typically 0.05 statistical significance). The rough average of the price spikes are 50% to 75% increases over the few months following October. Also note that the two largest storage shortfalls in 1996 and 2000 were followed by especially large price spikes. Also shown on the chart is the 2005 record hurricane season which had a price spike as well when natural gas production capacity was impacted in the gulf from hurricane Katrina and Rita. One final observation is the low point of storage in March. When that is especially low (which often follows a low October level of course) the price spike seems especially large. So this all makes good logical sense to me. Here is how I would summarize it:

1. The lack of natural gas supply to be able to keep up with winter demand on an instantaneous basis makes natural gas storage critical.
2. When the peak storage level in October is not high enough it suggests that winter supply may get tight...
3. Which leads to a winter price spike which can increase natural gas price by more than 50% and it peaks between November and February.
4. Hurricane or other major supply interruptions can cause a price spike for similar reasons. Hurricane price impacts are actually quite rare however, for a nice analysis see this.
5. This does not consider the impact of crude oil prices however which is also a strong effect especially evident last year. I will look at natural gas vs. crude oil more soon.

Natural Gas Price Forecast for 2009/2010 Winter

Based on this analysis, it is really too early to forecast for this next winter. However, the low storage inventory was just put in in March at 39% which is relatively high and April and May are trending quite high compared to historical norms. The EIA's June Outlook is projecting this October to be a record high inventory level. EIA is looking at absolute storage volume, not % of storage capacity which I looked at here. Plus, I think it is too early to make a call for this October. If you look at the above you will see that high levels at the low in March are not always predictive of high peak inventory levels in October. The other unknown as far as price is the effect of higher crude oil prices. There has been a growing gap as crude oil price has gone up in recent months and natural gas has not. I think this gap will be closed in the next 3-6 months. But that closing of the gap could be done through either natural gas price rising or crude oil price falling or some combination of the two. I will take a closer look at natural gas vs. crude oil in posts coming soon.

So bottom line, I think it is looking like natural gas won't have an inventory derived price spike this winter but it really is too early to tell. Hurricanes and crude oil still could cause natural gas prices to go up as well. The former is probably a long shot and the latter quite possible. I will update this forecast in a couple of months once natural gas storage inventories are further along.
...