9th February 2015
Analysis undertaken by GCA suggests that the current, and continued projected, fall in rig count may have as much as a 2 million barrels per day impact on US unconventional oil production.
Such a bold headline may seem to fly in the face of forecasts by other industry commentators and analysts. Indeed, the latest EIA forecast continues to show production growth through 2015 and 2016.
So, has GCA come to a different conclusion? Actually, no.
The analysis undertaken by GCA, focusing on the Bakken as an example for the whole of the US, confirms the potential for continued modest growth in production, although it also shows that, depending upon assumptions, many different credible “forecasts” are possible. It does show a big impact on production by the end of 2016, though – when compared to what production was likely to have been if there had been no oil price crash and the rig count had continued at previous levels.
Put another way, at the activity levels seen through Q3 2014, production in the major plays would have continued to drive strong production growth. With the cuts currently experienced and projecting into the possible landing zone, the reduced rig count is likely to be sufficient to hold activity constant for a while, possibly even still offering a modest increase.
If the production impact in the Bakken were to be extrapolated across other basins in the US, production at that point could be some 1.5-2 million barrels per day lower than where it might have been if nothing had changed. This is about the same amount as the current global oversupply according to OPEC, although other estimates place it somewhat lower. If demand remains in the doldrums, the imbalance in world supply-demand is likely to remain. However if the QE in Europe has an impact and Chinese demand also ticks up just slightly then oversupply could well come into balance within the next couple of years, allowing the oil price to move from speculative gyrations to being driven by fundamentals.
Discussion and Analysis
GCA has tracked activity and production output for the three major producing basins within the US; the Williston Basin, Permian Basin, and Eagle Ford shale of the Gulf Coast Basin (see also GCA’s Rig Count Monitor).
Rather than simply attempt to produce yet another forecast of production, using data from the Williston Basin (the Bakken), this article looks at several potential future activity and performance scenarios over the 2015 and 2016 period, and sees what sort of production levels are likely to result. It also examines the impact of a scenario currently being espoused by the largest operator in the basin, Continental Resources, whereby rigs continue drilling but completions are deferred.
In the week following the 30th January Baker Hughes rig count announcement and the announcement of (confirmation by) several large oil companies of budget cuts over the 2015-2017 period, the oil price rose US$10/Bbl (more than 20%) from its mid US$40s levels, albeit following something of a roller-coaster ride along the way.
A number of market commentators attributed the price rise to the rig count fall and budget cuts and that production would drop as a result, bringing supply and demand back into balance. While it is undoubtedly true that cutting investment will lower production (in due course), neither announcement is going to have any immediate effect on production. Indeed, neither factor was a surprise: the rig count drop was just the continuation of an already established trend and budget cuts were fully expected by the market. Thus the extent of the reaction, if indeed a reaction is what it is, must have more to do with financial trading than any sudden reappraisal of fundamentals.
A number of market analysts, including the United States Energy Information Administration (EIA), have produced forecasts or outlooks on US oil production for 2015 and 2016. In most cases the production outlook is for continued growth over this time window, or “flatlining” at worst. The EIA’s latest forecast, published on January 13, 2015, has U.S. oil production exiting 2015 some 350,000 Bopd higher than year-end 2014, and 2016 exiting a further 1,000,000 Bopd higher.
Putting the EIA and other public data into its proprietary field development model, GCA has looked at the assumptions required to meet such forecasts and test the extent to which production may react over the 2015-2016 period.
To develop the answer, GCA first looked to “history match” production in the basin against the BHI reported rig count as it has actually evolved. This was done in a number of steps:
Selecting a set of average “type curves” for production performance. This was based on decline curve analysis on historical productivity between 2012 and mid-year 2014
Modifying production performance over time for efficiency gains
Looking at the impact of well inventories (wells drilled but waiting on completion either intentionally or due to service/cash flow constraints)
The type curve was created by focusing on production performance from approximately 20 operators drilling in the four key counties that currently produce over 85% of the oil. This methodology was used to focus performance on those operators which are using best operational practices, as well as aligning the type curves with the anticipated production from the core area of the play.
The EIA reports monthly a view on the production additions in each basin from new wells, and the decline of production on existing wells. Over the course of 2014 the production added from new wells divided by the number of rigs active  (the “efficiency gains”) increased by more than 20%. This is actually a composite of more efficient drilling and completion practices, as well as better well production performance. The type curves and drilling activities were then adjusted similarly.
Lastly, the changes in reported production do not always exactly match the expectation from rig activity. There may be a number of reasons for this, but rigs, drilling and production hook-up do not necessarily proceed in a smooth and linear fashion. In particular, as rig counts are rising, there will tend to be a lag between new wells drilled and wells added to production. As of November 2014, in the Bakken there were more than 700 wells reported as being in inventory by the North Dakota Industrial Commission (drilled and, for the most part, not completed). GCA has assumed that in the short term the completion of some portion of these wells is adding to production.
The history match (and, as a result, the production forecast) based solely on the type curve and rig activity appears good up to late 2014, although GCA’s estimated production then drops below that reported when based solely on the rig count. This gap can reflect the above-mentioned assumption that operators would bring some well inventories into production to avoid the full cost of drilling new wells. As a result an assumption was included that there was additional completion of some 150 wells from inventory in late 2014/early 2015. This adds around 70,000 Bopd to the production output by April, 2015.
Having achieved an acceptable history match, GCA turned to production forecasting for 2015 and 2016. This was done as a scenario approach, initially testing independently the likely implication of different key activity parameters.
First tested was what might be considered a pessimistic scenario. In this case the rig count is assumed to fall to about 50% of the mid-2014 level by mid-2015, and then remain there. This drop follows the historical 2008/2009 decline observed in onshore rig count over a 6 month period. In this scenario no further efficiency gains were assumed after the start of 2015.
Oil production from the Bakken exited 2014 at approximately 1.2 million Bopd, slightly below the 1.3 million Bopd output forecast by the EIA as of February 2015. Despite the halving of rig count and removal of any assumption of continued operational and production efficiencies, production is calculated as still continuing to rise until around 2Q 2015, before falling back slightly and then holding fairly steady at around 1.1 million Bopd through the end of 2016. Thus, despite a sharp drop and then no pick-up in activity, by the end of 2016 there is only a 10% drop in production from the end of 2014.
A flat to modestly declining near-term production outlook might seem counter-intuitive in the face of such a sharp fall in rig activity. However, the reason becomes more apparent when looking at what might have been. Although slightly higher immediately before the price crash, around 180 rigs had been active in the basin in the late 2013 – early 2014 period. Production had been increasing steadily from a little over 500,000 Bopd at the start of 2012 to the 2014 exit of over 1.2 million Bopd, representing an annual growth rate of around 30%. Assuming no drop in oil price and a continuation of 180 rigs running (the “$100 case”), GCA estimates that Bakken production would have reached around 1.7 million barrels per day by the end of 2016.
As such, the halving of the rig count does in fact reflect a large drop in production - some 600,000 Bopd by the end of 2016 from what might otherwise have been the case if nothing had happened.
Additional sensitivities were then evaluated. The first was to assume that although drilling activity had halved, efficiency gains continue to be made in line with those seen in 2014. The second was to assume that while the rig count dropped by half it recovered to 140 rigs (approximately 75% of the pre-crash level) by the end of 2016, along with the efficiency gain.
As can be seen in the chart above, along with the $100 case, it is easy to conceive of many reasonable outcomes pretty much all of which result in flat to increasing production.
Reality suggests that the future will be some combination of all of the above circumstances. If the Bakken production over 2015 and 2016 is trended in the same manner as the EIA forecast US crude as a whole, Bakken production will experience a slight increase in late 2015, and will exit 2016 at 1.3 million Bopd, represents approximately 3% to 5% annual growth.
A further wrinkle in forecasts is also raised by a strategy espoused by Continental Resources Limited, the largest player in the Bakken. This scenario involves drilling but not completing wells straight away, seeking to defer flush production in the hope of capturing a price recovery or spike. This is actually the opposite of what may have been happening at the end of 2014/start of 2015 in depleting inventory, but can be considered as being analogous to players who are renting tankers to buy crude today, store it and then release it back into the market when prices are higher. The “storage” strategy may also explain in whole or part the sharp rise in oil price as players speculating in the market also seek to drive up the futures market and lock in early profits.
Depending upon the intensity and timing of any such deferral it may or may not have a significant impact by the end of 2016. It would however defer and possibly cause a meaningful drop in late 2015 production, creating a “saddle” before picking up when the inventory that has been built-up is completed and brought on production.
In summary, despite the expected sharp drop in capital expenditures and rig count, production over the next two years is not expected under most scenarios to drop from current levels, and may even continue to increase. However, this increase does reflect a sharp fall from where it might otherwise have been.
At the end of 2014 the Bakken represented approximately a quarter of US unconventional oil production. If that same percentage were to be maintained through 2016 and other basins were to behave similarly, and Bakken production is some 500,000 – 600,000 barrels per day lower than it might have been had there been no drop in oil price, then overall US unconventional oil production by the end of 2016 would be some 2 million barrels per day lower than where it might have been. It just so happens that this is about the same amount by which the markets are currently over-supplied. If demand can make up that up, something that can be achieved by only a little over 1% p.a. demand growth over the next two years, perhaps the recent bounce in oil price does start to reflect fundamentals.
If you are interested in following up on the contents of this article, or would like to discuss how it might be applied to affect your company, please contact your local GCA office, or in the US call +1.713.850.9955 and ask to speak to Bob George.
- GCA Oil & Gas Monitor
- Latin America
- North America
- Asia-Pacific & China
- Middle East
- Russia & Caspian
We're here to help
Europe / Africa / Middle East / Russia & Caspian
gaffney-cline & associates