Alexander Khurshudov: The US mightily accelerate gas activity
Moscow. I have been following changes in the American gas industry for almost 9 years, and, in my opinion, it is currently at its peak. Enthusiastic forecasts predict it will continue to grow for at least another 6 years, and I have decided to check how realistic the predictions are. As usual, I am writing for a wide audience, so I beg experts' pardon for the simplified representation of complex issues.
Last year the Americans produced 1,048 billion m3 of natural dry and associated petroleum gas. Not all of it reached consumers, 9.5% was pumped into productive oil reservoirs (mainly in Alaska, to increase oil recovery). After dehydrating and separating impurity gases, the EIA declared annual production of marketable gas in the amount of 862 bcm. Picture 1 shows production dynamics and its sources distributed by groups.
Gas production in old (conventional) fields has decreased 11.6 times over the last few years. This is not because the formations are completely depleted; if prices increase, it will still be possible to produce something there. However, the petroleum industry longs for new reserves like a drunkard for booze, but - alas - there aren't any in long-ago discovered fields.
Offshore production decreased 1.8 times, now it amounts to 31.2 bcm. Previously developed deposits are getting depleted, though it is possible to develop new deep-water fields off the East coast and around the shore of Alaska. However, the perspectives are not clear yet.
Gas production from low permeable (shale) formations has grown 3.85 times to 577.9 bcm. Annual growth has amounted to 50-55 bcm over the past 5 years. This is a lot. Almost 43% of shale gas is produced at the Marcellus and Utica fields, which have recently been combined into a single Appalachia basin.
Associated petroleum gas production has increased 3.3 times to 224 bcm over the last 8 years. Note that oil production has doubled, and associated gas production has tripled with a good surplus. Produced crude oil now contains 1.65 times more gas than it did 9 years ago. Let us keep this fact in mind, we will come back to it.
A question is how reliable these numbers are. They are reliable enough if we keep in mind some peculiarities of accounting process. According to the EIA data, for example, gas losses (flaring and dispersion in the atmosphere) are supposed to make up an insignificant part, 0.7%. It is doubtful. Here is another detail. Approximately 8.5% of processed gas is used, as they say, for their own needs - for heating lease facilities and gas processing plants, heating units and fueling compressors. This is a considerable amount, 72 bcm per year, but it is referred to the “consumption” section, although the gas industry uses the gas itself. In general, the EIA seems to interpret any doubts to the higher side; maybe that’s why BP gives another figure for US gas production, 832 bcm. While the initial data is the same.
- Reserves in place
Estimating proved reserves is the basis for petroleum geology, the basis for all subsequent decisions. At large, it includes determining deposits’ boundaries, rock porosity, oil and gas saturation and a possible recovery factor in case production is economic. Geological parameters vary in area and depth, technology improves, and economic efficiency depends on prices, which can rise and fall. Therefore, estimating reserves for large deposits is a very complicated thing; it requires 1.5-2 years of work and is done extremely rarely.
In 2011 the EIA estimated gas reserves in the first shale field Barnett Shale at 924 bcm. Over the next 8 years 393 bcm were produced (Fig. 2). The field has entered the final stage of development, drilling has stopped there and the accuracy of reserves estimation can be verified.
Gas production from shales has a valuable feature: repeated hydraulic fracturing (refrac) is effective here. On average, 3-4 operations are performed annually in each well. This enables to slow down production decline to slightly less than 10% per year. A simple extrapolation shows that another 190 bcm will be produced over the next 10 years, but recovering the remaining 341 bcm, if it ever happens, will take many decades. It turns out that the EIA reserves estimate is overrated by 20-25%, which is an acceptable error for a pioneer field. However, such estimates will always be too optimistic.
The reason is that EIA reports are based on data from gas producers. Every year they submit audit of reserves, hiring special appraisal firms. As a result of their well-paid work, in 2017, for example, the proved reserves of Barnett Shale grew by 68 bcm; with production steadily declining at the same time. More than once I have seen auditors overrated reserves, occasionally by several times, but usually by 10-20%. Yet I have never seen them underestimate.
Two reasons may contribute to US reserves increase - technological innovation and price increases. This is clearly seen in pic. 3.
The situation here is the same as with production: there is a decrease of 3.4 times in old (conventional) fields, 1.9 times decrease in offshore fields, while shale gas reserves increased 3.2 times, associated gas - 2.8 times. The 2014 peak was due to both price increase and massive drilling of wells with 2.5-3 km laterals and 25-30 frac stages. However, the reason for the 2017 peak is unclear to me. 18.7% price increase is a dubious basis for 46.7% increase in shale gas reserves, while technology potential is exhausted: increasing the number of hydraulic fracturing stages to 50 did not prove economic.
The increase in associated petroleum gas reserves is understandable, but it has a downside. Hard-to-recover oil reserves are always produced by depletion drive, it is impossible to displace oil with water, so reservoir pressure steadily reduces. At low pressure, associated gas begins to come out of oil right in the pores of the reservoir. It moves to the bottomhole along with oil, greatly complicating its filtration. Gas-oil ratio is grows gradually (this is the ratio of gas and oil production rates), and the well switches to pure gas and condensate. This process has been going on for several years now. As a result, the number of idle wells is growing; let’s take the Bakken field as an example (Fig. 4).
Note: of course, the increase in gas content in production is not the only factor contributing to wells shutdown. In 2010-2014 water breakthroughs were a serious complication. Last year, when oil prices rose to $70-80, operators put back into operation everything that showed some signs of activity. Finally, a part of wells is simply abandoned. For example, in 2018, 417 wells were abandoned in the Bakken. However, an increase in gas production is a deciding factor, because it is a sign of reservoir depletion.
Generally speaking, the US gas reserves are not easy to assess. Formally, they amount to 13.1 trillion m3, reserves life (at the current production of 1 Tcm / year) is 13 years. If the EIA estimate is overstated by 20%, then the reserves will last for 10.5 years.
- Large fields
A giant Marcellus field covering 246 thousand km2 is located in five states: Pennsylvania, West Virginia, Ohio, Kentucky and New York, but it is not developed in the latter, because hydraulic fracturing is banned there. Two years ago, the EIA issued a detailed report on this field; I will give a brief overview based on it.
The productive formation is buried at 600-2000 m and composed of tight fractured shale. Its average thickness is 15–20 m; in the most productive areas it reaches 150–180 m. The porosity of the rock varies between 2–5%, and its permeability is extremely low.
The average production rate of 52 thousand m3/day is very high for low-permeable formations. The reservoir here is disturbed by large faults, which greatly increase productivity, but can be the reason for an overly optimistic assessment of reserves.
These estimates really differ significantly. 8 years ago the EIA estimated technically recoverable reserves first at 11.6 Tcm, then at 4 Tcm. As of December 31, 2017, proved reserves were booked at 3.5 Tcm, then within a year, (due to re-estimation only) they increased by 47.2% (!!!). However, only the most productive formations, 20-60 m thick are being intensively drilled (Pic. 5). The total of 12.5 thousand wells have been drilled here. Last year production amounted to 201 bcm and continues to grow.
Last year, the Permian Basin oil field unexpectedly became the second biggest gas producer. 10 years ago, 150 thousand wells produced only 8 bcm/year, but now 180 thousand wells produce 76 bcm of gas. In 2017, the region’s gas reserves were estimated at 904 bcm, with an annual increase of 67%.
The natural gas saturation of oil here varies between 100-250 cm/t. In fact, due to the depletion of old wells, the gas-oil ratio increased to 1000 cm/t. After mass drilling in the shale, it gradually decreased to 1000 cm/t, but early last year it began to grow again (Pic. 6). Permian Basin seems to be repeating the history of the Bakken formation in a shorter time frame. I think gas production will continue to grow in the years to come, but I can’t say the same about oil.
Haynesville, located in Louisiana, Texas, and Arkansas, takes the third place in gas production. Its area is 23 thousand km2. Productive shales are buried at depths of 3-4.2 thousand m, which makes it somewhat difficult to drill and perform multi-stage hydraulic fracturing. Wells here are more expensive and take twice as much time to drill. But then the formations are much thicker - 60-90 m.
The first peak in gas production at Haynesville (72 billion m3) took place in 2012. At the time the EIA estimated its reserves at 836 bcm. Since then, about 5 thousand wells have been drilled here, production has declined, stabilized at the level of 50-60 bcm/year, but last year it reached 69 billion m3 and is now growing. The operators seem to have overcome the difficulties caused by the large depth of the deposit. As a result, proved reserves grew 2.76 times (!!!) last year only and now amount to 1017 bcm.
Other shale formations also produce a considerable amount of gas, but production there is gradually declining. Last year, Eagle Ford produced 45 bcm, Woodford - 29, Barnett - 27, Andarko - 25, Niobrara - 21 bcm. However, oil Bakken is steadily increasing gas production, it has now reached 15 bcm/year. Now let’s move on to gas consumption.
Over the last 8 years, gas consumption has grown by 24% and reached 846 billion m3 (Pic. 7).
Consumption for the gas industry own needs has grown by 26%. Residential gas consumption is growing weakly and takes up 16-19% of the total, because the US is a warm country. Industrial consumption has increased more significantly (by 21.8%) and, especially, consumption for power generation (by 44%).
The United States is making heavy use of the emerging opportunities to transfer energy sector from coal to more environmentally friendly gas fuel. Over the past 9 years, coal consumption has fallen by 34%. Near the Marcellus field alone 6.7 GW power plants have been built. Last year gas consumption for electric power increased to 301 billion m3, this year it keeps growing.
Nevertheless, domestic gas consumption opportunities are limited. It is worth mentioning that, despite the stimulation, transferring vehicles to gas is going slowly. During the last 9 years the consumption of vehicle gas fuel, though it grew 1.5 times, amounted to just 1.2 bcm/year, a little more than 0.1% of total production. Therefore, the United States is boosting the construction of gas export facilities.
For decades the United States has been an importer of natural gas, buying it from Canada and Mexico. However, in 2017, as a result of increased production and construction of LNG terminals, gas export exceeded import (Fig. 8). Last year, import amounted to 82, export to 102 bcm.
Over 70% of export went through pipelines to Canada and Mexico. 23% out of the 30.7 bcm of exported liquefied natural gas (LNG) went to South Korea, 17% again to Mexico, 11.5% to Japan, 8% to China, 5% to India, and 26 more countries received small volumes.
American LNG prices vary in the range of $140-250 per 1000 m3. Given the additional costs of regasification, they can only compete with local suppliers in Southeast Asia. As for Europe, spot prices fluctuated in the range of $170-360 per 1000 m3 last year; Gazprom’s average price was $246. However, now prices in the EU have fallen to $80. Therefore, Lithuania received only 2 gas carriers (108 million m3) from the United States and was satisfied with Russian gas.
The four existing LNG export terminals have a total capacity of 38 bcm (23 million tons) per year; this year it is planned to launch the fifth terminal and expand the existing ones, as a result the total capacity will double. Approximately 25 more LNG projects are awaiting review by the authorities. Such operating capacity requires long-term contracts and regular supplies, but it is not that simple. Picture 9 shows the dynamics of US LNG supplies to Europe; they clearly do not look too regular. In just 20 months, EU countries received 11.5 bcm, which is approximately 1.5% of the EU consumption.
American gas export is between the Devil and the deep blue sea. If domestic gas prices rise, it becomes unprofitable, but if domestic prices go down drastically, exports will set records, but producers will suffer losses and reduce production. The situation is risky for both sellers and buyers, so they are not inclined to entering into long-term contracts. However, the Polish national company PGNiG signed a contract for 23 years on very prudent conditions. The buyer acquires the rights to gas at the point of shipment, delivery by sea is at their expense. If gas in Europe is cheaper, they can redirect the shipment to other countries - to America or Asia. This year, the planned volume of gas (0.7 bcm/year) has already been shipped, with prices fluctuating strangely in a wide range of $ 112-274 per 1000 cm. Transportation and unloading will require another $70, but there is an opportunity to profit on the empty gas carriers. In short, the gas market is becoming a real market, where traders rule, while producers get leftovers.
- Recent news
Early winter last year happened to be cold in the USA, gas spot prices jumped to $145 per 1000 cm in November to producers’ great joy. Their happiness didn’t last long. A decline started in January, reaching $78.4 by August, a quarter lower than the previous year. At the same time, gas well drilling decreased by 11% (Pic. 10). The same is happening in oil fields, where the number of drilling rigs decreased by 18%, from 877 to 719.
They had to use suspended stock, that is wells previously drilled, but not completed. In Marcellus and Utica fields the number of such wells decreased by 251 within the 20 months of 2018-19. At such rates, there will be no suspended wells in stock in 3.5 years. For the second time, the rise in the gas industry is followed by a decline, and here, we can make a summary.
1. The perspectives of the US gas industry look quite good for the next 6-7 years. They are supported by available gas reserves, the rapid growth of domestic consumption and export infrastructure. I do not expect a rapid growth in gas production, but it may well exceed the current levels by 10-15%. It will be restricted by price fluctuation.
2. The key risk is an overly optimistic assessment of proved reserves. If it turns out that the reserves are really overestimated by 20-25%, then production will begin to gradually decline in 5-6 years’ time, and it will not return to the previous levels.
3. It seems to me that proved reserves of oil and gas (as well as other important minerals) are gradually losing their geological and economic importance. They are increasingly becoming an instrument for commercial, political and other manipulations.
Indeed, don’t they produce oil and gas at a loss? Do they shut down wells in case of a sharp drop in prices? Nothing of the kind. A good example here is Chesapeake Energy, which has produced 308 bcm of gas over 13 years with a net loss of $13.8 billion. Losses are covered by naive shareholders, creditors, selling assets, offset by income from new fields, there are also other ways. For example, Russia has so gracefully transformed its legislation that when oil price declines, taxes are reduced, and production does not fall, but grows, cheered by foreign consumers. Natural resources are the largest source of profit, and therefore are now ruthlessly exploited to full depletion.
4. If it is so, then the actual wells’ production becomes crucial. If production is high, actual reserves are also high. So far, we see that in all shale deposits only the central parts, fractured zones are developed, while the wells at flanges and slopes have much smaller production rates and deplete quickly. There, of course, the reserves are also calculated, but they are virtual and will be written-off in the future.
5. No matter when the decline in gas production begins, in 5 or 10 years’ time, it does not threaten America with any catastrophic consequences. For the global LNG market is growing by leaps and bounds. They won’t have to convert power plants back from gas to coal, export terminals will be easily reconstructed into import ones, and the world’s gas reserves are still 16.5 times larger than the US reserves.