Iowa combined temperatures

Iowa has 23 USHCN stations ranging from Albia to Washington, but it has no GISS stations on the list, which makes life a little simpler this week. I did continue the study as to how far back to run the temperature averages based on the current population estimates around the stations. The best value for Iowa came out to be only using the average of the last five years data, and that correlates with Nebraska and Texas, while Oklahoma was debatable. So I’ll continue that check for a couple more states, but it is clear that there is some consistency in correlation between the states.

What is interesting about the USHCN stations in Iowa is that most of them are in communities smaller than 10,000 and yet there is the same trend in temperature variation with population that holds true in other states. This is clearly indicating that the Goddard Institute for Space Studies (GISS) is in error to segregate its temperature data with a cut-off of 10,000 as the minimum number it considers for a station. Vide their list for Iowa . (Although they use none of these stations in the truncated number of US stations that they have shrunk down to). Note the following is just a sample from the list.



And here is the correlation, using the past 5-year average temperatures for the USHCN stations, using the TOBS data.

Average of the 2004-2009 temperatures for the Iowa USHCN stations plotted v local population.

Given that the same form of plot is evident in the other states, perhaps it is time for a gentle cough in their direction?

So let’s now look at how we got there and what other interesting things might exist in the data. Firstly the stations appear to be spread relatively evenly over the state:


Location of the USHCN stations in Iowa (USHCN)

The trend for the Time of Observation corrected data shows that temperatures have been rising in the state since 1885, at a rate of 0.5 deg F per century.



Iowa is 200 miles wide, and 310 miles long. It runs from roughly 89 deg W to 96.5 deg W, and from 40.5 deg N to 43.5 deg N. The center of the state is at Longitude 93 deg 23.1’ W, Latitude 41 deg 57.7’ N. The highest point is at 509 m, and the lowest at 146 m, with a mean elevation of 335 m. The average USHCN station is at 93.5 W, 42.2 N with an elevation of 319 m.

In that regard looking at the correlation of temperature with the geographical considerations in the state, first there is a strong correlation with latitude:



There is the correlation with longitude, which is again likely a by-product of the increasing elevation of the state as one moves west.


There is a much clearer correlation with elevation:


And so we come to the correlation of station population and temperature. As in the past I ran a secondary calculation looking at the change in regression coefficients as I increased the average number of years averaged to derive the temperature at each station. One is 2009, two is the average of 2008 and 2009 etc. The population data in this case all came from the citi-data set, and there were no missing cities to look for.


I therefore used the 5-year average data to derive the relationship of temperature with population.


It is interesting that the coefficient for the slope of the line is running relatively consistently with values that I have found for other states. (If I use a 5-year average the value for Nebraska is 0.808, and for Texas 0.875.)

There is still, however, a difference between the homogenized data shown in the USHCN record, and that of the original TOBS temperatures.

Gas prices and oil supply in light of EIA and OPEC monthly reports

I paid $50 to fill my tank at a gas station in Maine this morning, at a cost of almost $4 a gallon. When the Actress muttered some comment of protest, I told her that she had better get used to the price, because it is hard to see any normal reason for a decline in that price in the near future.

The EIA TWIP today was discussing the transportation fuel market this summer, and begins by noting:

Regular-grade gasoline retail prices, which averaged $2.76 per gallon last summer, are projected to average $3.86 per gallon during the 2011 driving season. The monthly average gasoline price is expected to peak at about $3.91 per gallon by mid-summer. Diesel fuel prices, which averaged $2.98 per gallon last summer, are projected to average $4.09 per gallon this summer. Weekly and daily national average prices can differ significantly from monthly and seasonal averages, and there are also significant differences across regions, with monthly average prices in some areas exceeding the national average price by 25 cents per gallon or more.

Well right now, before driving season starts, the price was $3.97 for regular – but the EIA have the “out” that this is after all Maine, which is at the end of the delivery line. Ah, well!! But I suspect that the EIA is still being a tad optimistic, and may regret that $0.25 error bar by the end of the season.

Their estimate, and the rationale for it are given in the new Short-term Energy and Summer Fuels Outlook with the price of West Texas Intermediate (WTI) at $112 (it has since fallen $5) . The EIA is expecting the market to tighten, based on the turmoil in the Middle East and North Africa, and “robust” growth of demand. But they only increase the anticipated average price of WTI to $106 this year, and $114 next. And in this I think that they are being rather too optimistic given the times. And that includes their estimate that the price of gasoline will still be below $4 (at $3.80 average) through the end of next year. (Though they do add a caveat that there is a 33% probability that prices could get over $4 on average this July).

And in an aside (since the topic today is mainly crude oil) it is worth noting relative to my post on the EIA World Gas Shale report that the EIA are projecting that the Henry Hub price for natural gas will remain around $4.10 per kcf in 2011 i.e. below the 2010 average, and it will only rise to $4.55 per kcf in 2012 – which doesn’t make those gas shale drilling balance sheets look any prettier.

The oil supply problem itself is sufficiently worrying. As with others they are still predicting a global increase in demand of 1.5 mbd this year, expecting that it will rise an additional 1.6 mbd in 2012. OPEC (whose daily barrel is currently at $117) expects that with the tragedy of the earthquake and tsunami in Japan, that there won’t be quite as much growth as previously expected, and thus are only anticipating a growth in demand of 1.4 mbd this year. As their April Monthly Oil Market Report notes, they do not expect countries outside of Japan to be affected, and thus they continue to anticipate a world economic growth of around 3.9%.

As I mentioned in an earlier post there were a number of Japanese refineries which were damaged, and the country which was refining about 4.5 mbd had an immediate drop to 3.1 mbd. That has now been partially restored as some refineries have increased production, and others have been repaired. However the country as a whole is still reported to be about 617 kbd short of the pre-earthquake figure. (And there are three coal-fired power plants Haramachi Tohoku, Kashima Ibaraki, and Hitachinaka Ibaraki that are still off-line and 9 of 210 hydro-electric plants were damaged. )

Nevertheless Middle Eastern suppliers stopped some of the shipments to Japan, and this may well be what is being seen as a short-term decline in demand. (OPEC saw a drop of around 0.5 mbd in tanker shipments in March). However OPEC anticipate that there will have to be substitution for the loss in Japanese nuclear power, since that cannot be restored or replaced with equivalent new nuclear power stations in less than several years. As a result they expect that the demand for oil as a replacement fuel (the loss could be made up by about 200 kbd of oil equivalent fuel) will increase later in the year.

OPEC expects that 0.6 mbd of the overall global increase in demand will be supplied by non-OPEC countries, with Brazil, the United States, Canada, Colombia and China increasing production, while the UK and Norway will show the greatest declines. That leaves the rest for them, and bearing in mind the loss from Libya, and other potential losses around MENA, though OPEC itself only expects to see demand for its oil increase about 0.4 mbd. (Which arithmetic doesn’t quite compute – but never mind – I am assuming that the rise of 0,4 mbd includes the offset to cover the losses in production within OPEC, and that with the 0.4 mbd OPEC increase, and the 0.6 mbd non-OPEC increase, that the world will only be 0.4 mbd short – which might come from NGL increases). Incidentally OPEC anticipates that Chinese demand will grow 0.5 mbd to 9.5 mbd.

There is an interesting comment in the OPEC report relating to the poor performance of natural gas prices (as I have been discussing).

Nevertheless, a sustained upward trend in HH natural gas prices may appear if there is a radical change in the US energy policy regarding nuclear production, which seems unlikely at present. According to Barclays, in order to rebalance the US natural gas market via higher demand, it would be necessary to shutter a large amount (13-26%) of total North American nuclear capacity.

Well I have to confess that is one answer that I hadn’t thought of applying in order to get the shale drilling companies off the hook.

OGPSS - The new EIA Shale gas report

I had intended starting my country-by-country more detailed analysis this week, but with the publishing of the EIA compendium on world gas shale deposits, and a little controversy I got into in comments elsewhere, I thought it worthwhile spending a post looking at the global prospects for gas shale. I am indebted to Art Berman for first bringing some of the problems to my attention, though I have since looked into the issue sufficiently to draw my own opinions, which this article reflects.

The EIA document (actually prepared on their behalf by ARI) begins by noting the importance of this new resource to the US Energy picture:

The development of shale gas plays has become a “game changer” for the U.S. natural gas market. The proliferation of activity into new shale plays has increased shale gas production in the United States from 0.39 trillion cubic feet in 2000 to 4.87 trillion cubic feet (tcf) in 2010, or 23 percent of U.S. dry gas production. Shale gas reserves have increased to about 60.6 trillion cubic feet by year-end 2009, when they comprised about 21 percent of overall U.S. natural gas reserves, now at the highest level since 1971.

Development of US shale deposits began to be aggressive around 2005, and “technically recoverable” gas resources in the USA are now considered to be 862 Tcf. With this discovery of this potential future domestic supply, it became wise to see what the equivalent potential was elsewhere, and thus the report which looked at some 32 countries, 48 shale basins, and 70 formations. The assessed basins (in red with an estimate, in yellow without) are shown in the map below.

Gas Shale resources evaluated by the EIA

Combined, the total resource base is then assessed to be 6,622 Tcf. Now it is important to draw a distinction at this point, because this is a resource. There is a considerable difference between a resource, which is something out there of value that may or may not be economic to develop, and a reserve, which is that fraction of the resource that is considered viable to develop. In the case of the USA for example, out of the 862 Tcf that makes up the resource, only 60 Tcf (6.9%) is considered to be viable as an addition to the US reserves.

Further (and a point I will get to later) gas from shale is somewhat more expensive to produce, relative to more conventional natural gas deposits. So, when the new volumes developed are put in context of the total natural gas resource available, while gas shales now provide a gain of 40% in the total volume of gas technically available, it may well be that the larger more conventionally available natural gas volumes may be less expensive to produce, and so will be developed preferentially in the next few years.

On the other hand, a domestically available reserve that does not require a nation to expend money on foreign fuels can have considerable benefit, even though it may be comparatively expensive. (And that is a judgment that brings in a lot of additional caveats, but it may very well drive development in countries without much other natural resource, and one thinks of countries such as Ukraine and Poland – where the alternate coal is becoming increasingly politically disfavored). Countries such as Russia and parts of the Middle East, where there are already large reserves of comparatively inexpensive natural gas were not considered in the study – so it is possible that the total volumes available long-term may be understated.

To derive the “technically recoverable” volumes the consultants looked at the volumes of free and absorbed gas available, after having decided the size of the shale deposit, and then used their best judgment as to how much of this could likely be recovered. In general they though it technically possible to recover between 20 and 30 percent, with some higher and some lower estimates within an additional 5%. They did not consider offshore deposits, nor critically, did they consider production costs or accessibility.

‘Free gas’ is gas that is trapped in the pore spaces of the shale. Free gas can be the dominant source of natural gas for the deeper shales.

‘Adsorbed gas’ is gas that adheres to the surface of the shale, primarily the organic matter of the shale, due to the forces of the chemical bonds in both the substrate and the gas that cause them to attract. Adsorbed gas can be the dominant source of natural gas for the shallower and higher organically rich shales.

I am not going to go into the details of the report as it pertains to individual countries, though this provides the bulk of the information that is provided within it. Rather I am going to include those resource values and the related discussion as I come to discuss the resources available to different countries. But as an illustration that not all the shale is likely to be productive, once could consider, for example, the drilling pattern for the Barnett Shale where most activity is concentrated in the area north of Fort Worth. And there are significant areas with very little activity at all.

Drilling activity in the Barnett shale (EIA)

However I will comment about the immediate significance of the resource, its size, and what I see as the short-term impact that these volumes will have on the global energy market. My sense is that they won’t make that much difference, once the initial hype and gasps over the size of the numbers has passed.

The reason for this is that this new resource is not cheap to develop. The initial cost comes in sinking the well, where once the initial vertical segment has been drilled, the driller must bend or deviate the drill until, when it reaches the formation with the gas in it, it is drilling sensibly horizontally. The driller must then hold the drill in the formation (which may be less than 50 ft thick) while the bit opens a hole that runs out perhaps two miles or more. The ability to sense where the bit is and to guide it along that path is a complex, expensive process, and you can’t take someone off the street and have them run a good well in a week. Over time there the number of qualified drillers who can do this has grown, so that in the week of April 8, Baker Hughes reports that of the 1782 rigs drilling in the USA some 50% were drilling for gas, and of the total 57% were drilling horizontal wells. It has taken some time for that level of expertise to develop.

Yet even with the right rig and crew, success is not assured, or even permission to drill the wells. One has also to get the right permits, and the right prospect. And here one can look to the map of the current producing wells in the Marcellus Shale in the Eastern U.S. to illustrate the point.

Wells in the Marcellus Shale (EIA )

You can see that the activity has concentrated much more in West Virginia than, say, New York State. This is partially due to the different compositions and inclinations of the state legislature in each state.

But even with the rigs, the crews the political support, and a well in a decent spot in the formation, success is not assured, or even the most likely outcome, because of the nature of the host rock. The major concern that I have had with the expectations regarding shale gas, relates to the rock in which it is found. Most natural gas deposits are found in relatively permeable rock, so that it is relatively easy for the gas not intersected by the well or the fractures to travel through the rock to those free spaces, and thus out of the well. The permeability of shale is, to be blunt, pathetic. That is, after all, why the expensive fracturing of the well, and the slick-water injection of proppants into those fractures is so critical. But the poor host rock permeability causes a fairly dramatic fall in production from gas shale wells after they are first brought on line. The easily accessible gas close to the fractures and the well makes its way out, and then the rest must struggle through increasingly thick layers of rock to reach the well. As a result there is a dramatic drop in production.

One of the wells that was cited in the comments debate I mentioned at the start of the post was the Day Kimball Hill #A1, which was the highest initial producer that Chesapeake ever drilled, at 12.97 mcf/day when it was brought on line. But:

The Day Kimball Hill #A1 is located in Southeast Tarrant County, Texas, and produced an average of 12.97 million cubic feet of natural gas per day in October 2009. Since shale gas wells decline sharply during the first few years, this Barnett Shale well has seen its production fall to 8.66 million cubic feet in November and 6.79 million in December.

That is a 47% decline in production in 3 months.

It was followed by White South #1H which came in at 17.8 mcf/day, last September. However, because of the low price of NG at the moment, the well has been partially shut-in to produce only 4 mcf/day since. Most shale gas wells don’t produce in this league. There are roughly 15,000 wells in the Barnett, and the wells in Arlington are an exception.

To join (the “monster” wells), a well’s output must average more than 8 million cubic feet of gas per day during its peak month.

Fewer than 1 in 1,000 Barnett wells has attained such a lofty yield for a month, which is enough gas to meet the heating and cooking needs for about 3,300 homes for a year, based on American Gas Association usage data.

While the Barnett Shale underlies more than 20 North Texas counties, the top “sweet spots” are in Tarrant and Johnson counties. All of the 35 biggest wells are in those counties, according to a new report by the Fort Worth-based Powell Barnett Shale Newsletter.

Day Kimball Hill site from Google Earth (Barnett Shale Drilling Activity)

Obviously, if you were a partner in these wells, you made your investment. There is a discussion of well economics which suggests that $5 per kcf is the minimum price (in 2005 costs) at which the wells can make money. But that assumes a 60% decline rate for well production over the first year. But in all the others, even as with these, decline rates from the initial numbers can be dramatically higher. And this is not just true for the Barnett shale . For example there is a report of a decline curve for the Haynesville shale that shows declines of up to 85% in the first year.

Haynesville shale decline curve (Haynesville Shale)

These high decline rates make it more difficult for the well to generate the return on investment that folk expect when they consider only the initial production volume.

The main problem that I see, in the short run, for the average shale gas well is that (as noted above) the producer really needs a price of better than $6 per kcf (and Art would argue higher) to make any money on the well. But there is a globally sufficient supply of natural gas more conventionally obtainable, and increasingly available as LNG at perhaps $4 a kcf delivered into the United States, that it is difficult to see the short-term viability of the industry.

On the other hand, with alternative sources of fuel failing to live up to various promises made for them, it may now be that natural gas is seen as the next hope for broad use vehicular fuel. Legislation has been introduced to encourage this use, and if this becomes more widely adopted then it may be that in this way the market may rise, and prices will hopefully rise with increased demand, to make the shale gas more viable sooner than I think.

Utah combined temperatures - revisited

When I first began this review of state temperatures I started in Missouri and moved west in subsequent posts. As a result Utah was the fourth state that I reviewed, and then, when the USHCN site also provided the Time of Observation corrected data, I looked at that data for Utah also. On April 4th Anthony Watts, at WUWT, wrote a post on the temperatures in Utah, based on the publication of a report by Weather Source, on temperatures in that state. As I have done, so they looked at populations in the state, but dividing population groups into three sizes, showed that temperature rise rates were different for the three groups.

Temperature trends for different station types in Utah (Weather Source)

The categorization of the station location as to whether it was urban, surrounded by agricultural activity (which will have some impact on temperatures) and those in natural wilderness or where the environs will have a low impact on the temperatures measured by the station. The study grouped the data in a different way than I do, and also only looked at the trends since 1948, while I look back over the data since 1895.

But primarily what I wanted to do today was to restructure the Utah posts so that they were in the same format as those of the later sites, where I included a couple of different plots and analysis not in the original posts. So this is a revision, and in the listing on the right side of the page it will replace the original two posts for Utah (though they will still be available through the references at the beginning of this article).

Utah has 40 stations that are not evenly distributed around the state, but clearly focus along a couple of highways.

USHCN stations in Utah

There is only one GISS station in Utah, and it is in Salt Lake City. (It has a full set of data for the period). So how does the GISS data compare with the USHCN homogenized data?


The average difference is that the GISS site is 4 degrees higher than the average USHCN site after the data for those sites has been homogenized.

Looking at the TOBS temperatures for the state over the 115 years:


This gives a temperature rise of 1.25 degrees per century, in contrast with the homogenized value of 2.55 degrees. (In comparison the Weather Source rates – calculated however only since 1948, give values of 4.2 degrees per century for urban locations, 2.7 degrees for agricultural sites, and 2 degrees per century for natural wilderness).

Looking at the Utah geography the state is 350 miles long and 270 miles wide. It runs from 109 deg W to 114 degW and from 37 to 42 deg N. The highest point is at 4,123 m and the lowest at 609 m. The average elevation is at 1,859 m. The average USHCN station is at an elevation of 1,604 m, and the GISS station is at 1,288 m. Given that there is a strong correlation with elevation (see below) it is not surprising that the GISS station reads high. The center of the state is at 111 deg 41.1” W, 39 deg 23.2”N. The average of the USHCN stations is at 111.6 deg W, 39.4 deg N.

In regard to the variation with latitude:


Note that the regression coefficient goes from 0.16 with homogenized data to 0.23 with the original TOBS data.

And with Longitude:


This tends to confirm that the true correlation is not with longitude but rather with elevation.


Turning to the relationship with population, it is not yet clear whether averaging the last 12 or 24 (or some value in between) for the number of years to correlate the temperature with is best. A quick check suggests that, of the three, a 12 year correlation is best:


And finally there is the difference between the homogenized and the TOBS data changes over the years. It is interesting to see that sudden upturn at the end.

Coal - the TV series

Well, thanks to the Engineer, I watched my first episode of “Coal” this afternoon. The show is on the Spike network with a significant part of the first episode setting up the story. The episode is called “The Master Mines”, and presumably relates to the skill of one of the continuous miner operators, who is able to produce 7 cuts – each 20 ft deep – during the course of his shift. The mine owners apparently need that production each shift (and there are two a day) if they are going to manage to keep the mine financially profitable. Well, needless to say, they don’t manage to get there, but they had to condense the action of about a month into one episode in order to generate enough suspense to keep the plot moving. (In contrast with the miners of Gold Rush– Alaska, which at the end of the season was doing that it seemed every couple of days).On the other hand these folks seriously know a lot more about what they are doing.

For those who have never been in a coal mine, or had any concept of what it is like, the series does give some impression of the mining operation, but it scurried through the introduction at such a pace that I am still not that sure of the layout of the mine underground. I am going to comment on some of the operations in the mine during the rest of the post, so that if you don’t want me to spoil the suspense, then you might want to go watch the show before reading on.

The Cobalt Coal mine is on Westchester Mountain in McDowell County, VA and two of the cast (?) took over the rights to mine the property, at the beginning of the season. The way that it was financed, with a $4 million investment, apparently requires that they must get enough cash flow from the coal (a high grade metallurgical grade lying in the Sewell seam that is sold for steel making) to cover the cost of operations even for the first month. There are 35 local miners who were hired to work the mine, and they have varying levels of expertise. The 7 cuts a day (is that one or two shifts?) translate into 40 truckloads of coal leaving the property for the wash plant, where the coal is cleaned of dirt (more of which anon).

The current working operation is an adit, i.e. a tunnel that runs straight into the mountain, mined from the coal, from the outside. It is only driven at seam height, and it is 600 ft long from the entrance (portal) to the working face. Those who are curious about such things as the layout at the face, and what the curtain is that they drive through, might get some help from the series of posts that I wrote on coal mining some time ago. The most relevant is probably the one about continuous miners, since that is the machine that they use.

The show really does not give the audience much idea as to how hard operating the continuous miner is. It used to be that the miner sat on the machine and steered it with joysticks from a cab on the side. But the machine is long, and the law says that a miner cannot go under unsupported roof. So taking the miner off the machine and controlling it remotely means that it can mine more before it has to pull back out of the tunnel it has dug, and move over to the next passage, so that the roof can be supported. The problem for the operator is that he can no longer see at all what is going on at the cutting drum. As the crew note, he has to listen and tell by sound, and the way that the machine is responding, whether the drum is cutting in coal, roof rock or in the floor. Every ton of rock that is mined is that much less coal produced, and that much more expense in cleaning and disposal. It is quite difficult to accurately steer and control the machine, and thus I suspect that the “apprentice” driver was much more skilled than the show let on. (Which raises a question about the night-shift operator, but that may be a series plot so I will not go there).

However the coal varies in height a little, and so the mine has to take some of the rock out of the floor (which is generally a softer shale) in order to give enough working height for the roofbolter – a machine that inserts what I believe are resin anchored roofbolts into the rock over the opening to hold it in place. This can mean that the machine has to cut as much as a foot deep into the underlying rock. The problem that occurs when it does (and generally it is done at the same time as the coal is being mined from the solid) is that the two get mixed together. Since the rock doesn’t burn it has to be separated from the coal, and that is the job at the coal preparation or wash plant. Here the mix is typically put into a fluid bath (simplistically) so that the coal floats, while the rock sinks to the bottom of the tank. The two are thus separated, and the cleaned coal can then be sold on the market. Incidentally when the bits are cutting rock they wear out much faster, needing costly replacement.

One of the “dramatic” bits of film was when part of the roof collapsed. It happens more often than you might think, and the roof has to be tested, even before it is bolted, and loose top broken down. Generally, however, miners use something a little more specialized than a geologic hammer – there are special tools and the modern ones are of fiber glass and quite long, so that the crew are well back from any rock that might fall on them. (Wasn’t always that way however).

I was glad to see the inspectors featured in the first episode, a safe mining industry is not only healthier it is also more productive, and we need a well qualified and powerful inspectorate to achieve that goal.

Well I may add more comments as the series progresses, it does give some picture of mining , though I have to admit reality was not usually as exciting as the series would make it, but, on the other hand, you don’t feel the muscle ache of working in that height when you’re watching it on tv – I still have knee problems from those days. (And yes I was underground when men died, and yes I was involved in roof falls). And in regard to back injuries, I led a research crew that worked on mining machine concepts (among other things) for four decades, all of us have bad backs.

OGPSS - The top 30 oil producers, a review

These posts have been going through the EIA list of the top oil producers in the world, over the past few weeks, I thought I might just review them collectively, but briefly, before starting to look at individual countries and oilfields. Even the posts that I have written recently have become out of date with new information (Russia increased production again in February by 20 kbd over January reaching 10.23 mbd) and then fell back to 10.2 mbd in March but at this stage, rather than focusing on such details, I am trying to generate a sense of the overall picture. It should also be recognized that I am just grabbing a snapshot of data, rather than the more detailed studies that look at the longer term, which folk such as Rembrandt, Rune and Euan provide. The simplest way to do this is to place my current estimates of production for the top 30 oil producers that I have reviewed in this series against the EIA estimate of their production in 2009.

Top 30 oil producing countries (those increasing production over 2009 are shown in red). (Click on the table to enlarge it)

It is significant to note that while Saudi Arabia was producing 8.05 mbd of crude in 2009, this has risen to 8.869 mbd on average for February as the Kingdom increased production to match the shortfalls in oil exports from Libya, inter alia. (With roughly 1.8 mbd in “other liquids” this takes total KSA production to 10.67 mbd and moves it back to the top of the League. However those numbers were from the March MOMR, which reports on February, In that report Libya was still being recorded as producing around 1.3 mbd). It is now reported that overall OPEC was not able to match the Libyan decline in March, falling about 350 kbd short, while KSA production has now reached 9 mbd, (10.8 including other liquids).

Contrary to President Obama’s recent remarks the EIA are anticipating a decline in US crude oil and liquids production over the next two years, part of which has been blamed on the change in GOM regulations. As a result it would be optimistic to anticipate much more than a US production of 8.3 mbd (and the EIA project it will be down to 8.2 mbd next year). It is unlikely that US production will increase beyond that point.

US crude and liquid fuels production – (EIA )

With China, Iran, and Canada holding relatively steady in the short term, this gives an updated total of 39.83 mbd for the top six, which is about 1.4 mbd higher than when I wrote the initial post back in February, but 500 kbd below the EIA estimate for their 2009 production. (While Russia and the KSA increased, the USA and Iran declined). Of these it is likely that only the KSA can continue to increase production much more.

In the second tier, Mexican production continues to fall, and was down to 2.556 mbd in February, with reports that it will now be an oil importer well before 2020. Exports have already fallen to 1.23 mbd, which does not bode well for customers. The United Arab Emirates (UAE) have, like the KSA, increased production to help out, though so far this has only been up to 2.394 mbd from 2.3 mbd for most of last year. (They also produce roughly another 500 kbd of other liquid fuels). By 2020 they should be able to produce up to 3.5 mbd. And in similar vein Kuwait, now producing at 2.368 mbd, up from 2,3 mbd. Kuwaiti plans are to reach 3.5 mbd by 2015, and be at 4 mbd by 2020.

The current political turmoil has even persuaded Venezuela to increase production, with OPEC reporting levels of 2.39 mbd for February, a gain of around 100 kbd. Though how long that is sustained depends on the success of the many investors that have been persuaded to invest in the Venezuelan oil sands.

In the third group Norway is declining, being now at just over 2 mbd, and even though it has just announced a major new discovery that will not come on line for at least 5 – 10 years, and in the meanwhile production will continue to fall. Norway needs more discoveries similar to this, however, to be able to sustain production levels extending into the future, since without them production will collapse.

Brazil was touted, by President Obama in his remarks about the Energy Blueprint last week, though the increasing volumes of oil that they will produce remain foreign to the United States, and though they will likely increase production up to around 4 mbd by 2020, rising domestic consumption may well take much of that increase.

Which brings us into the states that has some political turmoil. Iraq has been able to bring production back to around 2.64 mbd (according to OPEC) with the hope of reaching 3 mbd by the end of this year. At the moment about 1.2 mbd of this is exported. One of the great questions of the decade is just how close to a projected 10 mbd by 2020 that Iraq will be able to get. Sadly the continuing conflicts there, though reduced in scale, make it difficult for me to see much beyond 5 mbd by 2020.

Nigeria, which has had its own internal conflicts for some time, is going to the polls as I write this, and the expected winner is planning to overhaul the oil industry. However, if stability continues, then it might be possible to resurrect some of the older fields and perhaps increase overall production by some 350 kbd.

Algeria, which has had some turmoil, but may emerge from the ongoing protests without much change, is producing around 2 mbd of liquids. That has not changed as OPEC has moved to match the decline in volumes from Libya and other countries facing protests, and may reflect the current maximum that the country can produce. In the stability stakes I suspect that Algeria may survive without much change, although the plot I put up from Energy Export Databrowser does suggest that production may have peaked.

Algerian oil statistics (Energy Export Databrowser)

Angola is currently producing 1.7 mbd but may add some 650 kbd this year, for a total of 2.35 mbd. And that brings us to Libya, where the increased fighting, particularly over the oil refinery town of Ras Lanuf, makes it increasingly unlikely that the 1.7 mbd which came from Libya will be available again soon.

The United Kingdom is in significant decline, but recent moves to further tax the oil industry have made it possible that the decline may steepen. This because the new taxes proposed will likely reduce the profitability of the field developments proposed, discouraging their development. Recently production has run at 1.35 mbd of liquids, which is scheduled to drop to 1.3 mbd this year, and 0.94 mbdoe of natural gas, anticipated to fall to 0.85 mbdoe this year. The criticality of investment is shown in the projected production over the next 5 years, with the different colors showing the likelihood of success. Note that the grey of current production is declining at about 10%.

UK Projected Oil production (2011 UK Oil and Gas Activity Survey )

UK Projected Natural Gas production (2011 UK Oil and Gas Activity Survey )


Moving to the next tier down, Kazakhstan is now at 1.6 mbd and slowly increasing production toward a target of 3 mbd by 2020. Qatar is running at 1.4 mbd, but with almost 0.6 mbd of that in NGL. Indonesia is producing right around 1 mbd and may maintain that in the short term. It is being challenged in rank by Azerbaijan which has just incremented up to 1 mbd, a volume that is expected to continue to rise until it reaches about 1.25 mbd in 2014.

The tier that lies below 1 mbd starts with India, which is currently holding a production of around 878 kbd, and having to import increasing amounts of oil to meet demand. Given that the country also subsidizes the price, this is becoming an increasingly expensive consideration for the government. India is followed by Argentina, which is post peak and declined to 0.76 mbd most recently. Egypt is similarly declining, now to 660 kbd, but as one of the early nations to change under the most recent protests, and with the situation still somewhat fluid, it is difficult to predict how much the country will have both for itself, and for external customers, a year from now.

Oman will likely weather the current storms, and is also increasing oil production, to the point that it is moving up to pass India, with an Omani production of 863 kbd, some of which is tied to NGL production.

In the final four that produce more than 500 kbd Malaysia is barely maintaining production at 700 kbd, while Australia has fallen from 588 kbd to 540 kbd. Both are now being passed in production by Colombia, one of the “hotter” places for development at the moment, with production rising to possibly 920 kbd this year. Ecuador, which closes out the top 30, has recently increased production from 485 to 504 kbd.

That completes the top 30, and accounts for some 76.7 mbd of production. Those same countries back in 2009 were reported by the EIA as producing some 79.23 mbd of oil. Remember that world demand is anticipated to increase by somewhere between 1.4 and 1.6 mbd this year, and that of this list of 30 only 13 increased production, and the rest declined and the concern for the future becomes thus more clearly defined. (The difference between the two totals is partially explained by the loss in Libyan oil - we will see within the month how well OPEC covers that).

But it is not the overall production from the world that can be estimated that accurately, but by looking at individual countries and, in some cases, individual oilfields that we can get some better sense of what is to come. So the next step will be looking at these nations in more detail, in the weeks ahead.

Temperature correlation with population and the number of years to use

In the last few posts on state historic temperatures, I have changed the basis for the correlation between local station temperature and population from using an average of the station temperature over the whole period to just using the average of the past 30 years. However I chose a 30-year interval without any real basis for the choice. So I thought I would pause the state temperature reviews this week, to see what the best period would be to use for the temperature averages. Because the population data is relatively current (I am using the population given on the citi-data websites) the longer the time interval then the less likely an accurate correlation will be. On the other hand the likely scatter in individual station data might suggest that the longer the time interval the more accurate the correlation.

Note I was about half-way through writing this post when I read the Bishop’s post (Bishop Hill, not mine) on Lord Beddington’s comments on the UHI, specifically

Most stations are not affected by the urban heat island effect and there are well-established ways of taking the effect into account for stations that are (such as comparing temperatures on still and windy nights and excluding urban stations.).

I will have more comment on this statement later in the post, but the results that I am discussing demonstrably show, as I work through them, that the good Lord was wrong. (As I will discuss a little more below). But to return to answering the question first . . .

Starting with the last year of data (2009) I calculated the average temperature for each station and plotted this against the population data. (Though I did this for a number of states I will use the data from Minnesota as the initial state for illustrative purposes).

Temperature as a function of station population for Minnesota, using only the last recorded year temperature is reported.

Including a trendline on the plot, I then recorded the constant (35.714), the coefficient (0.7021) and the regression coefficient (0.16893) against the number of years used to calculate the average (in this case one). I then increased the number of years to two, by averaging 2008 and 2009 data, and plotted the curve again. I repeated this, incrementing the number of years in the average by one, until I was averaging 35 years in the plot.

Temperature as a function of station population for Minnesota, using an average of the last 35 years for the temperature value.

Beyond this point I stepped the number of years in 5-year increments out to 70 at which point I stopped. Then I plotted the three values in turn against the number of years in the average:

The influence of the number of years used in the average on the coefficient of the log of population for Minnesota

One thing to remember is that as more years are included, the added years are likely to be at a lower temperature, but if the temperature change is evenly distributed around the state, then the coefficient should remain the same. If, however, there is a change over time that differs from station to station, then the slope will change over time. Keeping that thought in mind, let’s now look at the change in the constant as the number of years is increased.

Variation in the base temperature in the correlation of temperature and population for Minnesota, as a function of the number of years used in determining the average temperature.

The third plot is the correlation coefficient, or r-squared value:

The variation in the regression coefficient (r-squared) determined for a plot of temperature with population for Minnesota, as a function of the number of years used to determine the average temperature.

The last plot suggests that (ignoring the one high value at year 22) the best selection for the number of years to include in the calculation is 14. However, before settling on this value I decided to check the values for other states.

The next state, moving backwards in the series, that I checked was Texas. When I had first made these calculations I had included the GISS station data as well as the USHCN station data in the calculation of population effect. One reason for this is that the GISS stations are, by and large, in the cities with the larger populations in each state. But as I have gone round the states in many states at least one of the stations that GISS uses has only data that starts in 1948. This also will affect the overall averages. Texas is not a state that has a good correlation between temperature and population, but, out of curiosity I ran the calculations described above for Texas both combining GISS and USHCN stations, and using only USHCN TOBS data. (I used Texas for this comparative analysis since it has a significant number of GISS stations). The correlation coefficient was significantly better when the GISS data was removed, and so I removed those values from all the correlations.

(This meant, inter alia, redoing the MN calculations, since this was originally calculated including the GISS values – the plots above, and from this point on, will not include the GISS values).

Texas values underwent a significant change where more than 10 years were included, and this is most clearly seen with the r-squared values.

The variation in the regression coefficient (r-squared) determined for a plot of temperature with population for Texas, as a function of the number of years used to determine the average temperature.

Part of the reason for this is that there were a couple of stations that only had data until about 2000, and when these were included they reduced the correlation. Without those values the correlation was higher, and interestingly, the coefficient was somewhat similar to that of Minnesota.

The influence of the number of years used in the average on the coefficient of the log of population for Texas

The next state that I looked at was Oklahoma, and ran the same procedure to see how changing the number of years included changed the trendline numbers. The plots that I obtained for the regression and population coefficients, using the same procedure, were as follows:

The variation in the regression coefficient (r-squared) determined for a plot of temperature with population for Oklahoma, as a function of the number of years used to determine the average temperature.

This would suggest either a dozen or twenty-three years should be used to compute the average temperature, looking at the two peaks. When the population coefficient curve is examined, below, it can be seen that at the point where the highest regression occurs, the value of the coefficient is around 0.65.

The influence of the number of years used in the average on the coefficient of the log of population for Oklahoma

The next state, moving steadily North from Oklahoma, is Kansas. This spoils the pattern since, although there are good correlations with latitude and elevation (the other two most significant contributors to varied temperature distribution around the state) there is not a good correlation with population. I did not do as many checks on the number of years for this state, as you can see below, but it is interesting that again there are peaks in the r-squared values at about 10 and 20 year intervals.

The variation in the regression coefficient (r-squared) determined for a plot of temperature with population for Kansas, as a function of the number of years used to determine the average temperature.

The correlation coefficient for the log population was about 0.1 in both cases.

Moving North again to Nebraska, one can run the same derivations and get:

The variation in the regression coefficient (r-squared) determined for a plot of temperature with population for Nebraska, as a function of the number of years used to determine the average temperature.

Which again suggests that either 10 or 20-22 years to derive an average would be better than 30.

The influence of the number of years used in the average on the coefficient of the log of population for Nebraska

Which again gives a coefficient value, in both cases, of between 0.6 and 0.7.

I am not going to belabor the point. In this series I have now looked at the temperature data for just over half of the contiguous states of the Union. It would appear that there is a statistical substantiation for the correlation between temperature and population around the measuring stations of those states of the form:

Temperature = state base temperature + a.logn (local population)

Where a is a constant with a value between 0.5 and 0.8.

I haven’t yet gone back and done the above calculation for all the states yet which would allow a more refined estimate of the value, nor have I yet done the correlation to the three parameters (latitude, elevation and population) together – though it is likely that the latitude and elevation influence the state base temperature, how they interact to control station temperature is not yet clear.

But for now I feel very comfortable stating that Lord Beddington’s statement is demonstrably, factually and terminologically inexact. (To misquote Sir Winston Churchill).