The Possibility of Water Being as Valuable as Oil

The Wild West is back—and seemingly unstoppable. Rebounding from one of most sobering downturns in its history, the oil industry is once again flying high on the wings of extraordinary productivity and efficiency. The good news? We have learned to do (much) more with less. Consider the Permian rig count, now 16 percent lower than “pre-crash” levels in 2014, but generating a whopping 200 percent more “new-well oil” per rig (~200 bpd/rig to 600 bpd/rig currently), according to the U.S. Energy Information Administration (EIA). And all this accomplished during the downturn.

The bad news? For the industry to achieve liftoff and stay airborne, it has some challenges to overcome. Consider “just” the Permian’s Delaware Basin. If you are operating in the region, your challenges may include a lack of takeaway capacity, infrastructure and personnel shortages, and stranded natural gas.

The good news? Each of these problems will be resolved by a free economy, entrepreneurial spirit and investment-seeking financiers.

But other challenges loom larger, and Mother Nature offers up a bit of a wild card. Demand for suitable water supply, and volumes and costs associated with managing water, are now shaping the behaviors of the industry. It will be interesting to look back at this decade to identify where, in fact, water becomes as valuable as oil. It may be closer than we think.


UIC RISK REPORTING HIGHLIGHTS TIMELY TRENDS

One of the hottest topics in the industry is the necessary role of UIC (Underground Injection Control) wells for oil and gas production. UIC capacity is emerging as a very real issue, which in turn, may become one of the strongest drivers for water reuse and recycling (in this case, as a means to avoid subsurface disposal).

Over the last three years, we’ve compiled research presented in our UIC Risk Report1—which included mining through tens of thousands of UIC-injection histories and assessing their risk profiles. We’ve observed several interesting takeaways.

#1 Seismic risk and formation (over) pressurization can both lead to capacity diminishment, but are caused by different risk factors.

Seismic Operational Risk
If injection programs are known to be associated with induced seismicity (generally low-magnitude events), persistent seismic activity levels can lead state regulators to enact new regulations or policies that restrict local area injection volumes. Recent examples of regulatory responses have included injection-rate caps, requirements for more frequent or rigorous pressure, volume or seismic monitoring, or in some cases, full shut in of specific injection intervals or formations.

Formation Health Operational Risk
The other cause of capacity diminishment relates to formation health and receptivity. Certain formations, which likely supported disposal operations for decades, are now showing signs of pressure increases or reduced injectability most likely brought about by years of prolonged, cumulative injection. Future capacity concerns may also stem from overlapping (intra-formation) injection and production activities.

While there is some overlap in risk factors between induced seismicity and formation health risk, many of the risk factors are different and should be evaluated independently. (See Figure 1.) spwm tworisksFig 1. - Two types of different risks

The bigger issue is that we are now experiencing signs of capacity restrictions at current productionlevels. Yet, leading operators predict substantial increases in oil productivity going forward, in part enabled by longer laterals and larger hydraulic fracturing (“frac”) jobs. However, if we frac with more water, and also produce three times as much oil, we’ll likely end up with three times as much produced water. So, today’s modest capacity constraints may soon become limiting factors as we scale to achieve our projections.

#2 Evaluating or predicting risk of seismicity is tricky, and there is much we don’t yet understand. For many seismic episodes or swarms, scientific opinions differ and continue to evolve on the causes of these events. But, just as the industry improved its seismic interpretations by processing old 2-D and 3-D seismic data with more sophisticated techniques, we are reprocessing and re-evaluating old earthquake-related seismic data to learn more about how to characterize and mitigate these events.

spwm seismiccauldronFig. 2 - Seismic-causes cauldron#3 A “seismic soup” of contributing causes can potentially contribute to seismicity. For an earthquake to occur, the area’s geology must possess certain features that predispose it to seismicity. Many oil and gas plays, lacking these features, have not experienced any seismicity in recent history.

However, in seismically predisposed locales, there are strong scientific correlations between recent earthquakes and a spectrum of possible causes. (See Figure 2.)

Many of these natural causes (such as compaction and settling) take place concurrently with oil-field operations, making it almost impossible to pinpoint a cause(s) with absolute specificity or accuracy.

#4 Induced seismicity isn’t restricted to high-pressure fluid movements. Fluids don’t have to migrate to an energized fault to bring about an earthquake. Transients in pore pressure, which can be instigated from fluid movement, can also cause earthquakes. Pressure perturbations can move great distances throughout formations and at high rates of speed. Thus, the location of the fault slip where an earthquake occurs may be at some distance from the originating cause. We rarely understand the exact pathways from the possible source of the earthquake to the epicenter or event itself. A case in point: Most of the earthquakes in Oklahoma took place at depths far greater than disposal-injection interval depths.

spwm seismicerrorofmarginFig. 3 - Seismic margins of error


#5 While we largely rely on factors like spatial and temporal proximity to help assess earthquake causes, a large margin of error is associated with the location of earthquake epicenters derived from then-available seismic monitoring systems.
Figure 3 shows the range of possible error margins the United States Geological Survey (USGS) assigned to individual epicenter locations in Texas and Oklahoma for earthquakes between 2000 and 2017.

The analysis shows that the true epicenter of a given event could be off the reported value by several miles (up or down, and side to side), based on the configuration and capabilities of the monitoring network and processing techniques used. Taken together, the unknowns (distance traversed, rate of energy transfer, unknown pathways, concurrent activities and inherent margin of error with monitoring networks) collectively make it quite challenging to pin down causation with absolute certainty. However, “circumstantial” data, including proximity to oil-field activities, should be sufficient to develop sound mitigation programs that minimize seismicity in many cases.

#6 Our data mining efforts and reviews of scientific literature suggest that injection-related induced seismicity is not limited to a high-pressure world alone. In looking at the 13 states we’ve assessed, a good portion of seismic events believed to be induced by injection also occur below low-pressure formations (such as parts of the Arbuckle formation in Oklahoma and the Ellenberger in Texas). One possible explanation is that lower pressure formations are more transmissive, allowing the pressure transients to more easily migrate from the injection point, through the transmissive formation(s) and to energized faults deep below. Deep energized faults in the pre-Cambrian basement are a known source of induced earthquakes. If this observed generalization about low-pressure formations and transmissivity proves true with further research, it illuminates a Catch-22 in the industry. For years, we assumed lower pressure injection formations were, in general, the best candidates for saltwater disposal (SWD) programs, with easier injectivity and lower pumping costs. But in seismic-prone areas, those low-pressure formations may instead work against us when it comes to induced seismicity.

We believe that ongoing research over the next decade will shed new light on the seismicity challenge and that mitigation programs will continue to improve to allow hydrocarbon production in safe and sustainable ways. Getting our arms around “Big Data” and enhanced analytic systems, and equipping regulatory agencies with next-generation systems, will play a central role in the long-term solution.


SHOULD WE BE CONCERNED ABOUT REDUCED INJECTION-WELL CAPACITY IN U.S. OIL AND GAS PLAYS?

Depending on where you operate, the answer is, unfortunately, a guarded “yes.” While seismicity is the more public-facing challenge with regards to SWD operations, in certain locales, we believe constrained disposal well injection capacity—from formations that have provided many years of injection service and now show signs of pressure regime changes—will be of even greater operational and financial relevance to industry.

This problem is manageable—but at a cost. A receptive disposal well can always be found, but it may be a good distance away from where the fluid is generated. This drives up logistics and transportation costs, often the greatest cost components of water-management programs.

Formation Health Indicators
spwm macroanalysisFig. 4 - Macro-level pressure-volume analysisThroughout our regional assessments in the study, we evaluate both surface-pressure trends (obtained from surface pressure-gauge measurements) and bottomhole pressure (BHP) trends (estimated from standardized equations incorporating dynamic surface pressures, estimated friction-pressure contribution and hydrostatic-pressure contributions). While there can be a wide margin of error in pinpointing true BHPs for a specific well from surface and operating conditions, by applying consistent methods across the entire UIC well population at various intervals in time, macro trends (changes in aggregate pressure ranges across a formation) become evident. The findings in Figure 4 come from analyzing all available recent performance data (many tens of thousands ofUIC wells) for the most recent three-year period available for the five states profiled.2

Basic Pressure Volume (P-V) Analysis 101
If a formation behaved as a truly “closed” vessel, expected behaviors include:
A. If volumes increase (determined by rate of injection), we expect pressure to increase.
B. If volumes decrease, we expect pressures to decrease.

However,
C. What if volumes decrease, but pressures continue to increase? 
Could pressure increases, in the face of decreasing volumes, be a result of the cumulative effects of injection?

D. What if volumes increase, but pressures decrease?
Could that suggest that pathways allowing pressure dissipation are present? Should lost-circulation zones be explored? Is the formation transmissive, facilitating fluid or pressure movement?

Assessing the degree to which these four distinct P-V relationships are occurring by basin, formation and injection intervals sheds light on formation health and disposal risk factors as a component of a fuller risk assessment. A simpler way of looking at macro-formation pressure trends (only) is to compare total volume increases to total pressure increases. For all wells injecting into a given formation within county boundaries, determine:

• Percent of volumes injected associated with BHP increases over the same period (A+C)
• Percent of volumes injected associated with volume-rate increases over the three-year period (A+D)

The difference between these two determinations suggests the degree to which pressure increases are outpacing volume increases or are occurring even though injection rates have declined. While this analysis doesn’t translate to specific formation behaviors associated with a specific well, it does provide broad visibility into the long-term capacity trends of our UIC systems in the most prolific oil and gas basins. Interestingly, regions likely to have higher levels of natural seismicity, such as parts of Oklahoma and Colorado, tend to show higher volumes associated with behavior D, suggesting transmissive pressure pathways may be present. 


spwm pvtrend2Fig. 5 - Volume increases versus BHP increases (M Bbbls) spwm pvtrend3Fig. 6 - Volume increases versus BHP increases (%)  spwm pvtrend1Fig. 7 - Texas Plays 

Across all Texas formations, over 3.1 billion barrels of fluid injected in 2016 were associated with county formations having increasing bottomhole pressures* 2014 - 2016.

For nearly one-third that volume (~900 million barrels), the pressure increases occurred despite declining injection volumes, suggesting future capacity may be constrained.

*As derived, not measured

As a percent of all volumes analyzed, New Mexico showed the most dramatic differential of the five states in this analysis, with BHP increases (82 percent of formation volumes in 2017) substantially outpacing volume/rate increases (14 percent of formation volumes in 2017).

(The New Mexico analysis is more current, 2015 - 2017, than other states analyzed, 2014 - 2016).


In Texas, no region appears to be immune from potential capacity changes going forward for long-used formations.

Throughout most counties in Texas, average rates of injection declined substantially during the period we studied (starting with the downturn in 2014). The expected behavior, therefore, would have been related pressure reductions, not increases.

However, every Texas play also has optionality, with many receptive UIC formations available and without noteworthy risk factors. We believe the landscape will become more competitive, however, for receptive formations going forward.


THE ABILITY TO PLAN AND PROJECT HAS BECOME ESSENTIAL

Large oil and gas operators wishing to stay ahead of disposal-related risks are beginning to apply increasing geological, geophysical and seismic scrutiny to disposal formations.spwm uiclayersFig. 8 - UIC layers

Small to mid-size independents, who may not have the luxury of a dedicated geological and geophysical team for disposal formation analysis, will apply additional diligence and planning to assure long-term operations continuity. But, they may rely more heavily on third parties and published studies for guidance, which was one of the underlying drivers for developing the UIC Risk Report.

Using either approach, the intent is to recognize trending risk factors early in the process, such that alternative strategies can be adopted to sidestep potential capacity constraints or mitigate seismicity in some regions. Determining optionality early in the process is also important to keeping costs under control.

Applying Insights to the Local Level
Once regions of interest are identified, these same investigative processes should be applied at the local level. By loading relevant injection and completion data, cleansing it, converting to BHPs and plotting derived BHP against injection depth, a dynamic injection pressure gradient analysis can be conducted to visualize the pressure regimes in a local area. Findings warranting further investigation could include pressure increases (limiting injectability), outlier high pressures approaching frac gradients, high utilization of regulatory permit-assigned maximum surface pressures or unaccounted pressure losses.

County & Depth Comparisons of Dynamic Injection Pressure Gradients: One Way to Visualize Pressure Trends
We developed methods to display data for visualization of pressurization trends. In Figure 9, we compare pressure regimes for 26 counties in Central Oklahoma at the county level by 1,000-foot depth intervals, with higher-pressure gradients shown in red. This viewpoint gives broad, high-level insight into an area that also has been characterized by injection-induced seismicity in some counties.

spwm injectionpressuregradientsFig. 9 - Central Oklahoma dynamic injection pressure gradientsOklahoma is renowned for the extent and depth of the prolific Arbuckle formation, much of which is underpressured. Much of Oklahoma’s historical seismic activity has occurred in regions underlain by the Arbuckle, where extensive and deep low-pressure intervals (shown in green in Figure 9) are present. Note also that quite a few counties in Central Oklahoma are characterized by shallow intervals with relatively high average injection pressure gradients, which overlay low-moderate pressure regimes below. This particular trend is more typical of the trends observed in the Permian in Texas.


INJECTION PRESSURE GRADIENT MAPPING: ARBUCKLE FORMATION, OKLAHOMA 2015

Mapping these same data points allows for visualization of the geographic extent of pressure trends. For example, as Figure 10 illustrates, not all of the Arbuckle is strongly underpressured; there is significant geographic variability within the formation.spwm ipgmappingFig. 10 - IPG Mapping Arbuckle Formation 

We can explore further at a local level by plotting the dynamic injection pressure gradient against depth just for wells in our area of interest. (See Figure 11.)

• Lower or underpressured well-injection intervals will appear toward or below the hydrostatic pressure gradient line (0.465 psi/foot, normal hydrostatic pressure for 100K TDS formations).
• Higher pressure intervals of interest will dominate the right-hand side.
• Moderate pressures will typically reside in the middle, somewhere in the .5-.7 psi/ft range.

The benefit of this type of analysis is that wells from multiple formations in the region can be compared to help identify optimal injection targets in the local area. Intervals showing both high- and low-pressure regimes can be readily visualized and compared for UIC targeting and risk profiling.spwm threechartsFig. 11 - Three charts of dynamic operating injection pressure

We rely on these different analytic methods in our report, specifically so that a trend of interest can be cross-checked from different views to confirm the trend and ascertain the geographic extent of pressure trends.

Disclaimer: Key qualifying information such as injection volumes and well counts is also essential, as statistics derived from very small data sets will not be as meaningful as analysis performed on many hundreds of wells in a given region or formation.

For any analysis, the combination of risk factors will substantially vary and must be assessed at the local level for an actionable water-management plan. However, region-scale analysis can give us visibility into high-level trends, which can help guide and prioritize further assessments at the local level. Factors including geologic history, maturity, seismic profile, operating history, oil and gas activity levels, and the cumulative effects of long-term operations are among contributing factors that should be included in a local assessment, if possible.


THE PATH AHEAD IS MANAGEABLE

In the case of disposal constraints, beneficial outcomes are likely to include water-use moderation and more recycling by operators, accelerated adoption of pipelines and more utilization of rights of way, a stable market for haulers, and it can be hoped, increasing interest in developing beneficial reuse of produced water by other industry sectors.

In conclusion, we might debate whether water will indeed be as valuable as oil going forward, but there is no question that our ability to manage natural resources in the oil field has to become more sophisticated. Assurance of operations continuity is going to require long-term planning and strategic (and possibly creative) thinking.

Since we can’t manage what we don’t measure, self-monitoring will play an increasing role in our disposal operations, as will more sophisticated data management and modeling.

Returning to our Wild West theme... if you haven’t started the process yet, it’s a great time to saddle up!

 

Authored by Laura Capper

 


References

1 EnergyMakers Advisory Group Industry Report: U.S. Class II Subsurface Infrastructure: Injection and Seismicity Operational Risk Factors and Mitigation Strategies

2 Years 2014-2016 for Colorado, Texas, Oklahoma and North Dakota, and 2015-2017 for New Mexico