Preface: this article does not make any investment recommendation. Its purpose is to provide product market information relevant for evaluation of investment in frac sand mining companies such as EMES, HCLP and SLCA. All facts and data as of 8/25/2014.
FRAC SAND OVERVIEW
Frac sand is a type of silica sand (also referred to as silica or quartz) that meets API standards for size, roundness and sphericity and is used in hydraulic fracturing. Fracking is a well stimulation method that involves pumping of a special fluid under high pressure to generate cracks in target rock formation. This allows oil and gas to flow from the rock into the well. Water and dedicated substance called "proppant" make up about 98% of the fluid. Proppant needs to be strong enough to withstand pressures of up to 6,000 psi and at the same time to be made of round and uniform pieces - to allow the flow of petroleum. Certain types of silica sand meet these requirements and are available at a fraction of cost of synthetic proppants such as ceramics and resin-coated sand. That's why silica sand dominates proppant market, with 80% of all proppant sales in 2012 (according to Freedonia Group), despite worse technical performance for certain applications.
As a result of geological processes millions of years ago, the type of sand best suited for fracking - Northern White - is in abundant supply within 50 feet of ground level in western Wisconsin, eastern Minnesota, with some deposits also in north-eastern Iowa and northern Illinois. This map by Minnesota Environmental Quality Board shows first encountered bedrock; sandstone is marked red or solid pink (the black rectangle highlights the area where WI, MN and IA border each other):
For a detailed description of fracking process, frac sand and sand mining, if inclined to do so, please refer to: (I) Silica Sand Mining in Wisconsin, 2012, by WI Department of Natural Resources, and (II) Report on Silica Sand, 2013, by MN Environmental Quality Board - both available for free online.
FRAC SAND MARKET IN THE RECENT PAST
Rapid demand growth and supply shortage resulted in skyrocketing price
Frac sand usage has increased exponentially over the last decade, driven by the exploration of North American shale basins that require hydraulic fracturing. Prior to the shale oil and gas boom which started around 2006, silica sand was produced primarily for manufacturing (glassmaking, foundry etc.) and construction. Some sand was used for well stimulation in conventional basins, but on a much smaller scale. In 1996 (the earliest USGS data available) total U.S. production of silica sand for all uses was 25.5MM tons, with frac sand accounting for 6% of that. In 2012, the volume of frac sand alone (31.1MM tons) far surpassed the 1996 total sand production level, with frac sand accounting for 61.9% of all silica sand production.
Several features specific to fracking explain why this growth was and still is so fast:
- in the tight shale plays, fracking is typically performed in horizontal wells, which are usually longer than the vertical part of well (e.g. a typical Marcellus Shale well is drilled 5,000-9,000 feet vertically and up to 10,000 feet horizontally); longer wells mean longer distance from surface to fractures that sand needs to cover;
- increasing number of lateral legs stemming from one vertical well - reduces drilling costs making incremental sand usage cheaper;
- increasing number of fracking stages per well - for example, number of horizontal frack stages is expected to increase 19% in 2014, 19% in 2015, and 16% in 2016. This results in increased sand usage per well - from 2,500 tons in 2012 to 5,000 in 2013-2014, with some wells using up to 8,000 tons (according to SLCA Q1 2014 conference call).
Price of frac sand has increased almost as quickly as volume - 86% since 2010. This happened because sand mining industry could not increase production quickly enough to meet the demand. Technology was not the main factor that slowed down the capacity expansion, since it only takes about 6 months to set up a major (2.5MM ton/year) sand production facility (according to EMES Q2 2014 conference call). Nor was it scarcity of sand (please see the next section for details).
The key inhibitor to sand mining expansion was environmental concerns. Since 2010, when frac sand mining in Wisconsin and Minnesota became noticeable, the local public got utterly concerned about variety of possible adverse effects from sand mining, including silica dust, water contamination, noise, destruction of picturesque hills, damage to tourism, increased traffic and even overall well-being.
Political battles raged in Wisconsin and Minnesota since 2011, with activist groups petitioning governors to ban the industry permanently. Among most important recent developments were (i) failure in April 2014 of Wisconsin state legislators to pass a bill that would limit counties' and townships' authority to regulate water and air quality, and (ii) refusal in April 2014 of Minnesota governor to impose a 2-year moratorium on frac sand mining, leaving the issue to local authorities. Out of 31 counties or townships in WI, MN and IA that have frac sand mining facilities as of May 2014, ten had sand mining moratorium at some point between 2011 and 2014.
Even in this environment, the number of sand-mining and processing facilities in Wisconsin increased from 5 in 2009 to 135 in 2014. But overall, such a regulatory climate was not very conducive to major long-term capital allocation decisions. Apparently, this uncertainty about frac sand supply helped sand producers negotiate higher prices and longer contracts (including take-or-pay contracts) with fracking operators who were understandably willing to pay more and commit for longer to secure the vital component.
FRAC SAND MARKET FORECAST
Rapid capacity expansion is highly probable, sand price stabilizing, looming collapse of shale drilling long-term
For the next 3-5 years, frac sand price is not likely to keep growing at the same pace as in 2009-2014, primarily due to imminent capacity expansion. On the demand side, there is increasing evidence that shale development, effectively the only consumer of frac sand, is economically unsustainable. Obviously, a slowdown, let alone a halt, in shale drilling, will ruin frac sand stocks. However, getting the timing right is crucial, and I cannot make a recommendation without thorough research specifically of shale drilling industry, which will require a separate study.
1. Supply - new facilities coming online will at least stabilize the price
Mining moratoriums are in the past
As discussed above, regulatory uncertainty was an important deterrent to sand capacity expansion since 2010. But things are changing. Looking at the moratorium statistics combined, it's hard not to notice that: (i) over half of places with frac sand deposits didn't even discuss a moratorium; (ii) 18 out of 31 places with frac sand facilities didn't even discuss a moratorium; (iii) of 27 moratoriums that were enacted, 25 expired or were repealed; (iv) 63% of moratoriums ended at least a year ago (according to county and state websites and local newspapers, for comprehensive list, please see Exhibit 1).
In a typical moratorium lifecycle for a given county/township, things started with concerned citizens, public discussion that usually resulted in county/town board enacting a moratorium for anywhere between 6 and 12 months. A committee was appointed with the task of assessing the impact of sand mining industry and formulating policy proposals. In a few rare cases, original moratoriums were extended (or, on the opposite, repealed before deadline). But typically, the committee did submit a report on time, the local authority adopted or amended its non-metal mining ordinance and the moratorium expired.
The newly-adopted mining ordinances differ by locality, but here are highlights of one of them from Allamakee county in Iowa. It is one of the most recent ones (adopted in June 2014) and it was hailed as very onerous - so it gives a good idea of the high end of restrictions:
- conditional use permit required;
- mine application will include several surveys, studies and plans covering road impact, geological factors, wildlife habitat, soil types and depths, wetlands, archaeological and cultural resources erosion, dust and site reclamation;
- all property owners and residents within two miles of the proposed site would be notified and given a chance to comment on the proposal, with public hearing;
- reclamation efforts to begin during mining operations, posting of reclamation assurance bonds;
- specific requirements for assuring air and water quality. For example, mining would not be allowed within 1,000 feet of a sinkhole or other karst landscape feature, within a mile of a stream or river or within 40 feet of the ground water table;
- mining sites cannot exceed 40 acres and must be at least five miles apart and at least 2,000 feet from churches, schools or residences.
To fully understand how restrictive this is, a mining expert would be in order. But from a common sense perspective, it looks like all these rules amount to higher up-front and compliance costs and longer lead time - which, if anything, is one of the makings of industry consolidation, but not a fundamental impediment to production. Next, it was the uncertainty, not costs in themselves that inhibited sand capacity expansion (let's not forget those ROE figures of 42 and 56% for EMES and HCLP). And, finally, to put things in perspective, Allamakee county doesn't (and didn't) have industrial frac sand mining to begin with, meaning that there was probably no industry influence in the policy-making process...
One important development to watch will be the release of moratorium report in Trempealeau county, WI in mid-September 2014. With this county having the highest concentration of frac sand facilities in the nation (12 operational and 14 in development, as of May 2014), its population probably can claim the most damage/concerns about the industry. The moratorium is confirmed to expire on 9/01/2014, and this expiration is strong enough signal in itself. But in September the study committee will release its report to the county board, and this report might set a landmark precedent for other counties and states. Given the county's high profile, the findings and policy recommendations from this report will give a good guidance on the future of sand mining compliance.
On a related note, the statistics of air permits are indicative as well. Between 5/20/2014 (the earliest data available) and 8/15/2014, Wisconsin Department of Natural Resources has approved all five applications for sand mining and processing facilities (according to its website) - which, by the way, gives a glimpse of the speed of looming capacity expansion. To conclude, it appears that regulatory uncertainty around sand mining is no longer a factor.
Still a lot of sand available
Although a reliable public estimate of fracking-suitable sand deposits is not available and not all sand deposits are equally profitable (or profitable at all, for that matter) due to differences in overburden thickness (how far sand is from the surface), nevertheless just by looking at the map it is hard to make a case that there is no more sand left in WI and MN (maps from University of Wisconsin-Stevens Point and MN Dept of Natural Resources). Railroad access in an important consideration for mining economics, since a producer's ability to deliver sand directly to a frack site at competitive cost is predicated on minimizing load/unload operations. And it looks like Wisconsin railroad network is dense enough in sand-bearing areas to at least not rule out new mines and plants:
Even with the possibility of higher transportation costs in incremental facilities due to their relative remoteness from railroads (compared to established ones), there is still enough gross margin at the current price point to keep new facilities profitable, as the next section will demonstrate.
New sand mining and processing facilities would still be highly profitable
Using Hi-Crush Partners (NYSE:HCLP) as a benchmark (because of its pure-play exposure to frac sand it allows for cleanest estimates), and considering the straightforward nature of sand mining technology, it would be helpful to estimate profitability of new sand mining operations. HCLP breakeven price (inferred from income statements by dividing total costs excluding taxes by volume produced, adjusted for sales of purchased sand) and average realized sand price since 2012 can give some reference points:
To estimate facility cost, we can use three recently built facilities as a benchmark:
Assuming the facility would be built by a group with comparable access to capital, engineering and construction expertise (or one of these very operators), and factoring in increased environmental compliance costs and inflation, we can conservatively use a cost of $75MM for a 2MM tons/year facility. With these inputs, it turns out that additional sand mining facilities are still highly profitable under reasonable range of breakeven and selling prices:
Granted, the tolerance window in terms of breakeven price is not very wide. This might deter inexperienced entrants who will inherently have higher per-unit costs. But the established players such as HCLP are not only comfortably within the range, but have a potential to decrease their breakeven price by spreading higher production volume over same of slightly larger fixed costs.
Summary: new supply is imminent and is coming online
As a validation of regulatory feasibility and persistent economical attractiveness of frac sand mining, 15 new facilities were under construction in Wisconsin as of May 2014, with 15 more permitted and 16 more proposed:
As a result of actual and expected increase in supply, frac sand operators are finding themselves in a more favorable negotiating position than they used to be in 2010-2013. This is already visible in the evolution of sand supply contracts. HCLP and U.S. Silica (NYSE:SLCA) don't disclose this information, but Emerge Energy's (NYSE:EMES) dynamics are pretty telling (based on EMES SEC filings and Q1 2014 conference call):
- revenues under efforts-based contracts (whereby "customer is relieved of his obligation to buy if unable to use or resell sand due to adverse market conditions" - EMES filings) increased from 9% to 27% of total between 2012 and 2013;
- efforts-based contracts share of all contracted volume increased from 18% to "over half" between 12/31/2013 and 3/31/2014;
- efforts-based contracts share of newly-added contracted volume increased from 42% to 65-70% between Q1 and Q2 2014.
2. Demand - undeniably present, but fragile
It is hard to argue with unrelenting growth in historical frac sand consumption. And the multi-year supply contracts that EMES and its peers keep securing for the millions of tons of sand coming online, speak for themselves. However, it will be instructive to explore how sustainable this demand is and what signs to look for.
Questionable economics of shale development makes it unpredictable
There is an increasing debate about economic sustainability of shale development. Googling "shale oil companies losing money" brings numerous links worth looking at. Here is a brief compilation from the most in-depth and factual sources: Post Carbon Institute, Labyrinth Consulting Services and U.S. Energy Information Administration.
The main problem with shale plays is their extremely high decline rates, unlike anything seen in conventional basins (decline rate measures decrease of in well production in a given year):
Decline rates this high necessitate continuous new drilling to maintain production level to meet investors' targets. Considering how costly shale drilling and completion is ($7-8MM per well compared to $0.2-3MM per conventional well, according to Breaking Energy, Marcellus Drilling News, Siemens and MIT), this creates significant funding need for exploration and production companies. To make matters worse, far from all shale wells are profitable, for example only 29% of wells in Eagle Ford field were commercial as of December 2013, due to failure or inability to identify core areas prior to leasing and drilling, according to Labyrinth Consulting Services. And finally, some basins are already so densely drilled that they are close to full development (e.g. 5.5 years of drilling left for Eagle Ford play (Ibid)), meaning that incremental drilling will almost inevitably happen in worse areas, resulting in lower initial production and yielding lower returns, if any.
These factors put substantial strain on oil and gas companies' balance sheets. Bloomberg's study of 61 shale drillers in May 2014 showed that their debt has doubled over the last four years while revenue has gained just 5.6%. In a wider set of 127 major oil and gas companies, during the year ending 3/31/2014 cashflow from operations totaled $568bn, compared to uses of cash of $677bn. The shortfall was filled through a $106bn net increase in debt and $73bn from sales of assets. The gap between cash from operations and major uses of cash has widened from $18bn in 2010 to $100-120bn during the past three years, according to U.S. Energy Information Administration. In a different setting, increasing CAPEX would be a welcome news, as a signal of growth, but seeing businesses borrow to build more wells that will lose over 90% of their production in 4 years (if they produce anything at all) is a little disturbing.
To summarize, there is mounting evidence that skyrocketing increase in shale drilling (and consequently in frac sand consumption) was fueled not entirely, if at all, by attractive economics, but rather by cheap money and yield-hungry fixed income investors. There is no question that some E&P companies might be making profits in shale fields, but collectively the industry is losing money and getting itself deeper in debt. Unless a sharp rise in oil and gas price puts money back into the system (at the expense of the rest of the economy), at some point this will have to come to a screeching halt, effectively destroying shale-related stocks and especially frac sand ones.
This is a separate and potentially highly profitable short theme, certainly deserving a thorough review.
Natural gas price is dangerously close to breakeven
Petroleum exploration and production is only economical while energy prices stay above certain level, known as breakeven price. For shale plays this level is higher than for conventional ones, because of costlier drilling and extraction techniques. Its exact value, obviously, is unique for every well, but there are several average estimates. For natural gas, they range between $2.88 and $5.22 per MMbtu, with one as high as $8.76/MMbtu (according to ITG Investment research, Arthur Berman, MIT Energy Initiative, Spectra Energy, Enercom Consulting). Excluding this last outlier, the average of five published estimates is $3.98/MMbtu. This price point is very close to the current gas price (as of 8/14/2014):
Theoretically, a sustained drop in gas price to under breakeven level would render shale gas drilling uneconomical, drastically reducing fracking volume and consequently demand for frac sand. But in reality, the line between drilling for oil and gas is blurry, since more than half of wells produce both oil and gas (based on well data by EIA, compiled in Exhibit 2) and each of the seven major shale plays produces both oil and gas. While Marcellus is almost entirely gas-oriented volume-wise (97.7% of new well production on energy-equivalent basis) and so is Haynesville (97.4% of new well production), oil makes up non-trivial part of their production on dollar-value basis (9.3% of new well production in Marcellus, 10.5% in Haynesville).
This oil component will provide extra cushion for drilling economics well below the estimated gas breakeven price even in the predominantly gas basins. Nevertheless, although a drop in gas price will not halt drilling (and by extension fracking) entirely, there is no question that it will make it less profitable, excluding marginal locations/wells from fracking pool, thus reducing demand for frac sand.
The next-closest basin by gas concentration is Utica with its 78% of new well production coming from gas (on energy-equivalent basis), but in dollar terms it is dominated by oil (55.2% of new well production). The remaining four basins are almost entirely oil-producing. So it is only Marcellus and Haynesville that will be affected by decrease in gas price.
It is hard to reasonably quantify the reduction of fracking activity without access to wells statistics by basin, including frac sand usage, and such information is not publicly available. But using shale rig count as a crude proxy for fracking activity (assuming same utilization, drilling speed and fracking intensity across all 7 major basins), gas-dominated Marcellus and Haynesville account for 10.9% of national shale rig count, as of July 2014, according to Baker Hughes. In an extreme hypothetical case when all drilling in these two basins stopped, demand for frac sand would decrease by 10.9%. But with the above-mentioned oil cushion and different breakeven prices for each well, the actual impact would be smaller.
Oil price - main fracking driver
Publicly-available estimates of breakeven shale oil price also vary substantially, yielding an average value of $70.25/bbl, based on publications and comments by ITG Investment Research, Post Carbon Institute, U.S. EIA, Wood Mackenzie, Baker Hughes, Reuters, Bloomberg and The Economist. Oil price is currently roughly 30% above this threshold, looking like there is plenty of safety for ongoing drilling and fracking operations. However, considering the previous discussion of the unsustainable nature of shale exploration and the financial and operational strain than oil companies are under, the true breakeven price might be much closer to the market. And, unlike that of natural gas, price of oil is the primary input for fracking economics, as oil accounts for 83.9% of all new-well production across all shale plays in dollar terms (a drop in oil price would shift this split but not by much - e.g. with 20% drop in price oil will still account for 80% of production value). As EIA publishes per-rig numbers for new-well production instead of absolute ones, rig-weighted average is used to arrive at the 83.9% ratio. Using ongoing production (with absolute numbers), the share of oil is 75.3% of all production from 7 major plays in dollar terms.
Since economic feasibility is not binary but gradual and breakeven prices are different for different wells, basins and operators, a decline in oil price would not halt all shale fracking entirely, but will exclude marginal wells from the fracking pool, thus reducing demand for frac sand. In an extreme hypothetical case where all shale drilling (and consequently fracking) would cease, demand for frac sand will probably go back to pre-shale levels of 2004-2005 at 4MM tons, making it a 10-fold drop from the 40.8MM tons of 2013 and bringing ruin upon all frac sand stocks.
Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.