[Editor's note: This story originally appeared in the January edition of E&P. The article below is an extended version with additional charts from Rystad Energy. Subscribe to E&P magazine here.]
Link to AnYiCheng
A Rystad Energy analysis has found little evidence that lower crush strength proppants appreciably change production decline curves in the Midland Basin. Yet in the Delaware Basin, brown sand appears to increase the decline rate when analyzing operators in aggregate.
Proppant is required to withstand high temperatures and pressures downhole to allow hydrocarbons to flow to the wellbore. Proppant had sometimes been made of synthetic ceramics, but operators have determined this is rarely worth the cost. Historically, operators believed that only Northern white sand, mined in Wisconsin and neighboring states, fits the strength and shape requirements for most shale wells. However, brown sand mines in Central Texas have long served frac sand to the Permian Basin even before . The downturn forced operators to cut costs and, as a result, brown sand volumes increased in and . In late , in-basin sand mines began to come online in the Permian, cannibalizing demand for Northern white sand and imported brown sand.
FIGURE 1.In recent years, operators have pumped an increasing amount of proppant per stimulated foot, while simultaneously lengthening laterals. To mitigate the resultant costs and logistical difficulties, operators and service companies alike have switched to lower cost in-basin sands. Yet concerns remain whether the new design profile of wells&#;which require massive amounts of lower crush strength sand&#;will result in steeper production decline.
In-basin sand adoption in the Permian is at approximately 80%, which is among the highest of all plays in the U.S. The Delaware has deeper formations with higher downhole pressures, whereas the Midland has shallower formations with lower downhole pressures. In-basin sand adoption on the Midland side of the Permian Basin has historically been higher relative to the Delaware due to this very reason. There are about 20 active in-basin sand mines serving the Permian.
Methods
When analyzing a well&#;s performance with available data, the most important factors to consider are the well&#;s location, operator and completion date as well as the mass of the proppant used and lateral length. Different acreage locations have different production potential; therefore, the effects of proppant will differ. Operators will apply varying techniques, which may have a significant effect on production but will not appear in publically available data. Similarly, completion techniques change over time; thus, models must keep the completion date of analyzed wells in mind.
The total amount of proppant and the length of laterals will both increase production. These are mathematically simple to control, and many linear models find that controlling for one eliminates the need to control for the other. Lateral length can thus be controlled by dividing production by lateral length. Because well production rises slightly less than linearly with lateral length, this method is not perfect, but it should hold when examining small groups of wells with roughly similar lateral lengths.
Data on the sand type used are sourced from self-reported numbers in the FracFocus Chemical Disclosure Registry. Coverage is not complete, and there is some bias for wells where Rystad does not know which sand type has been used. Rystad also has made an effort to fill in data gaps by utilizing other sources when possible.
Midland Basin results
In previous studies, Rystad Energy did not find a statistical relationship between production decline in the Midland and the type of sand used. As seen in Figure 2, initial results show that for and horizontal wells, brown sand had little effect on well curves in the Midland Basin.
Rystad identified 68 brown sand fracs and 126 white sand fracs operated by Operator A in and 262 wells operated in the same region by other operators in .
FIGURE 3.Operator B&#;s brown and white sand wells performed similarly in terms of barrels of oil produced per lateral foot, though brown sand fracs displayed slightly lower IP per lateral foot. However, both outperformed other Midland wells in the area. Rystad cannot conclude from this analysis alone that lower crush strength sands had no effect on decline rates, as the data do not have visibility over a variety of factors. Nevertheless, it does appear that Operator B has come to this conclusion itself, as the company appears to have continued to use brown sand in fracs and was one of a handful of companies that publically announced its usage of lower cost sand.
Operator C presents another interesting, counter-intuitive case study. Rystad positively identified that 37 of its wells in were fracked with white sand, 11 with brown sand and 37 with a mix of white and brown sand. Rystad compared these wells to 173 wells fracked by other operators near Operator C&#;s acreage.
FIGURE 4.Operator C&#;s production profiles revealed that its white sand wells significantly underperformed its brown sand wells, its mixed sand wells and offset wells completed by other operators. One major reason is the lower proppant loading, averaging 20% fewer pounds of proppant per foot than its brown and mixed sand wells. This may demonstrate why operators choose lower cost proppant, which means operators can use more proppant for less money, usually leading to higher production.
Delaware Basin results
Intuitively, it might be expected that brown sands have a greater effect in the Delaware Basin, as treatment pressures are higher than in the Midland Basin.
In and , Rystad saw higher IP but faster decline for brown sand wells.
FIGURE 5. FIGURE 6.Figure 7, a scatterplot of all wells in the Delaware Basin, shows a slightly positive correlation between proppant mass and decline ratio. In this case, a higher decline ratio means slower production decline. This positive relationship does not appear to exist for brown sand wells. As in the Midland Basin, more insight can be gained by examining operators that used both brown and white sand in the Delaware Basin.
FIGURE 7.Rystad positively identified Operator D as having 25 fracs using brown sand and 16 using white sand in . Operator D also exhibited similar production profiles for white and brown wells. Rystad could not compare these against wells fracked by other operators in the vicinity, as Operator D has a large acreage position.
FIGURE 8.Conclusions
Rystad found little evidence that the usage of lower crush strength proppants appreciably changes production decline curves in the Midland Basin, and the behavior of operators that have used both Northern white sand and Texas brown sand within the same area suggests that these operators have come to the same conclusion.
In the Delaware Basin, Rystad found evidence that, in aggregate, brown sand increases the decline rate. However, Rystad did not see this pattern when examining individual operators. Instead, Rystad saw that for one operator, brown sand wells performed very similarly to white sand wells. For another operator, Rystad again saw lower IP but not a faster decline.
An expanded statistical study of the aggregate effect of sand type in the Delaware Basin would be informative, controlling for downspacing, depth, treatment pressure, frac fluid type and other variables. An economic analysis, contextualizing the difference in performance between white sand and brown sand and taking into account the difference in costs in the form of a net present value or internal rate of return analysis would be important next steps in understanding the effect of sand type in the Delaware Basin.
(All charts are courtesy of Rystad Energy)
Read E&P's other January cover stories:
Stimulating Future
Key Frac Ingredients Moving In Different Directions
Where shale producers in Texas are getting their proppant has turned out to be one of the sector&#;s most significant market transformations since the onset of the oil industry&#;s downturn.
The 4-year movement toward using regional sands from within the state has been a major economic win for the operators leading it. The production results are encouraging. Supply chains are more efficient. Yet, there remain long-term concerns.
Distinguished by their roundness and high-crush strengths, northern white sands from the US Midwest had served as the ubiquitous proppant of choice for much of the unconventional revolution&#;supplying 75% of the sector&#;s frac sand in . Then crude prices collapsed.
Northern white sand accounts for just over 40% of today&#;s market. Nearly a quarter of the supply now comes from within the Permian Basin thanks to more than a dozen new in-basin sand mines that opened within the past year.
The additional supplies have forced the premium northern white sand suppliers in the US Midwest into price parity with the regional sands in Texas. However, without the need for cross-country rail shipments from Wisconsin or other northern states, the all-in cost of the Texas alternatives is actually about 60% less&#;amounting to a reduction of $40-$50 per ton.
Previously, operators avoided the use of these regionals sands due to their lower quality. The fear was that their more brittle grains tend to compact quicker, cutting off hydrocarbon flow from within the fractured reservoir. But for the most intense horizontal well designs that use 40 to 50 million lbs of sand, the arbitrage can amount to more than $1 million of freed up capital per well.
&#;The way I look at it is, if you can save that much money, then it doesn&#;t take very long until you&#;ve already funded a new well,&#; said Scott Forbes, a proppant expert with the consultancy Wood Mackenzie.
Forbes has been visiting the in-basin mines in Texas and speaking with buyers for the past year. In terms of production issues, he has heard of isolated cases where improper sand cleaning operations at some of the new mines may have contributed to poor well completions.
Overall though, the sentiment Forbes has picked up on from the shale patch is that the gains outweigh any minimal losses seen so far. Companies that were the first to switch to regional sands &#;are pretty well sold on it,&#; he shared, adding that their gains are creating pressure on the holdouts to follow suit.
The big sand shift started early on in the downturn as producers in Texas sought economic refuge by &#;debundling&#; their reliance on service companies and sourced brown sands (i.e., Brady or Hickory Sands) directly from mines in the central part of the state. The second act&#;the arrival of the in-basin operations&#;is solidifying the transition sooner than many anticipated.
For more proppant sandinformation, please contact us. We will provide professional answers.
Kent Syverson, chair of the department of geology at the University of Wisconsin, has been tracking the proppant sector for several years and said the development represents the biggest shakeup in the US sand market since its inception.
He explained that the ongoing demand destruction of northern white sand is being exacerbated by a recent confluence of factors in the Permian: producers exhausting annual budgets, limited pipeline capacity slowing down drilling, and the arrival of the new mines&#;some of which operators have taken full or partial ownership of.
&#;All of those things kind of came together in August and started to dramatically reduce the demand for northern white sand&#;and its price,&#; he said. &#;It&#;s amazing how fast it has set in.&#;
Rystad Energy issued a market analysis in December that concluded that &#;the adoption of in-basin sand in the Permian has evolved at a faster rate than we initially expected.&#; The report cited a &#;long list&#; of more in-basin mines still under construction that may yield an annual increase of 25% in US sand supply by the end of .
Forbes also forecasts that uptake of regional sands in Texas will continue and is seeing the increased competition pushing suppliers to offer incentives such as first-time customer discounts. He also expects in-basin mines to become a sectorwide norm. Oklahoma and the DJ Basin in Colorado and Wyoming are seen as the up-and-comers in the sand play.
&#;Anywhere that has activity today, if there&#;s not already local sand presence, there are companies looking to find local sources,&#; he said. &#;Eventually, I think there will not be a play out there that doesn&#;t have local sand capabilities.&#;
Trains for trucks: long supply chains for sand deliveries from the US Midwest to fracturing operations in Texas are being replaced with long lines of trucks. Sources: Annushka Peck and Supreme Sand HaulersOutside of the savings, a main pillar of the shale sector&#;s confidence with in-basin and brown sands is that they are almost as durable as the northern white sands.
This is a measure of crush strength, which for the northern white sands (depending on mesh size) can exceed 10,000 psi; viewed as strong enough to withstand the forces of fracture closure stress in most US shale plays. A number of the in-basin and brown sands top 8,000 psi ratings; viewed as good enough to cover all but the deepest and highest pressure formations.
However, comparable crush strengths have not fully eroded the original concerns for all the experts.
Jennifer Miskimins, an associate department head of petroleum engineering at Colorado School of Mines, warned against overweighting crush strengths since the shape of the sand grains is known to be an undermining factor.
She explained that northern white sands have earned their premium status thanks to being blown back and forth over long geologic stretches, giving them their robust, spherical shape. In-basin and brown sands have not benefited from nature in the same way, and generally have crystal structures with sharper corners.
&#;With that angularity, the sand inherently packs tighter and maybe has a little less crush strength, so it tends to have lower conductivity right off the bat,&#; Miskimins said, adding that this degradation, or compaction, grows worse over time as the angles of the crystals &#;cut into each other and break pieces apart.&#;
Looking for these effects within the data can be a difficult task since shale producers often need a year or more of production to establish key factors such as drainage areas or estimated ultimate recoveries. Equally challenging is the ability to isolate performance drivers when horizontal well designs pre-shift used sand volumes that were several times smaller than what is used today.
&#;So things can look really good early on,&#; Miskimins points out. &#;The question is, did you over-capitalize because you put in 100 [fracture] stages?&#;
The shape, not the color, is among the most important factors in sand used as proppant. Rounder is better, while more angular grains tend to break down faster. Source: Core Labs, Stim Labs.
Core Labs&#; Stim Lab Division is one of the companies that shale producers are asking to help assess the quality of in-basin sands. Robert Duenckel, vice president of Stim Lab, has published extensive research on proppants and said the firm&#;s analysis shows that in-basin sands skew toward a lesser quality but are often comparable to Brady (or brown) sands&#;but none proved to be as good as white sands.
&#;It&#;s quite a range of performance,&#; he said of the in-basin sands. &#;But we haven&#;t seen any that perform better than what we typically see for the Brady sands.&#;
Most of the analysis on in-basin sands is based on physical property testing, which Duenckel said offers a &#;snapshot&#; of a sand&#;s physical characteristics such as roundness and crush strength. To get a more complete understanding, he emphasized that conductivity tests are essential.
A conductivity test would reveal how a sand holds up under formation stresses and temperatures. However, these tests are more expensive than the property tests and thus fewer are run.
This has left Stim Lab with a limited data set to issue nuanced conclusions on in-basin sand conductivity. But an ongoing study will use computer models to shed more light, including the long-term degradation of in-basin sands and potential impacts on well performance.
Despite the current constraints, Duenckel said early results indicate that in-basin sand quality will be a bigger issue for higher-quality reservoirs (i.e., those with the most resource potential), which &#;should logically require a better proppant with higher conductivity.&#; In these scenarios, he expects degradation is more likely to &#;eventually show up in terms of poor performance.&#;
Though more data is needed to fully assess the performance of in-basin sands, there is enough to scrutinize the surge of brown sand usage.
After combing through public data on more than 20,000 brown sand wells in Texas, researchers with Rystad found no clear correlation to negative production. All of the examined wells were in either the Permian Basin or Eagle Ford Shale and had been producing for at least 12 months.
&#;If you look at the type curves, they have been going further up in the graph,&#; said Thomas Jacob, a senior analyst at Rystad. The low-price factor coupled with the lack of negative signals &#;is the reason why you&#;re seeing a lot operators switch to in-basin sand or the local sand that&#;s available.&#;
What the data points to is that better acreage positions and improved fracture designs are bigger influencers or production than the sand being used. The results of the study were published this year during SPE&#;s Annual Technical Conference and Exhibition (SPE ).
The researchers acknowledged that the public data might not illuminate the entire production picture. But if there is an economic impact on the long-tail of a well&#;s productive life, the early payout is providing comfort. &#;Everyone would prefer to have money in hand now, not later,&#; Jacob said.
Despite its diminishing standing as the pre-eminent frac sand supplier, the US Midwest will still hold plenty of buyers in the major basins that are closer to home: Bakken, Marcellus, and Utica Shales.
But stability is about the best the northern white sand market can hope for, according to most projections that forecast output to maintain current levels for the next few years.
That could change though. Several mines shut down over the summer, while others reduced output&#;leaving open the option to shutter.
&#;And we&#;re probably not done,&#; Syverson said, noting that some companies may proactively mobilize their equipment to set up new roots in Texas and Oklahoma, or sell it to the new mine operators already there. Whenever the machinery is cannibalized, &#;at that point, the capacity is never going to come back here in Wisconsin.&#;
With so much new supply poised to enter the market, frac sand prices are expected to remain flat into the near future. Source: Rystad Energy
For more information, please visit frac proppant.