Evaluation of the Capacity for Water Recycling for Colorado Oil and Gas Extraction Operations

A major logistical concern with requiring oil and gas operators to utilize produced water volumes for HF operations is whether there are enough volumes of produced water to satisfy the needs of HF activity. The following sections provide critical analyses of the existing regulatory data in order to answer this question. In addition to looking just at water use and water production, the analyses below also provide insight into the response of oil production to increased HF water volumes.

Hydraulic Fracturing Water Volumes

Statewide

Beginning with statewide data, plotting annual sums of water consumption alongside  wastewater production provides a broad overview of the relationship between water in vs water out. The chart below in Figure 2 shows annual volumes of water consumption for hydraulically fracturing (HF) oil and gas wells in Colorado compared to the counts of completed HF operations. As of 2022 the volume of freshwater injected during HF operations has more than doubled (tripled in 2018) while the number of HF events has decreased 60% since 2013. Figure 3 shows that average volumes of water used for HF completions has steadily increased since 2013 as well. Maximum injection volumes have varied widely on the other hand. In sections below, plots (Figures 10-12) show that the drastic increases in HF water volumes have had a minimal impact on increasing oil production.

Figure 2. Water use for Hydraulic Fracturing operations and Annual Counts of Hydraulic Fracturing Operations. The plots sum the total water volume consumed for HF operations each year and overlays the count of HF completions reported by FracFocus each year.

Figure 3. Water Consumption per HF Completion. The plot shows the average and maximum water volumes reported consumed by FracFocus for individual HF completion events in Colorado, 2013-2022.

Keeping in mind that water production data is under reported, particularly for 2020 through 2022, the data in Figure 4 shows that, with the exception of 2018 when HF operations consumed over 15B gallons of water, volumes of produced water have remained consistently greater than HF water consumption volumes. While the capacity may exist statewide, the transportation of produced water long distances typically requires tanker trucks, which degrade local and regional air quality as the miles traveled increase. While a statewide water sharing pact may be possible, the pipeline infrastructure required to accomplish this goal is likely not feasible at this point in time.

Figure 4. Statewide data plot of water volumes used for HF and Produced Water Volumes. The plot presents statewide sums of water consumed for HF operations and overlays the annual sums of produced water volumes reported to COGCC by operators. Of note, the sizeable reduction in HF water consumption in 2020 was due to the COVID-19 market decline.

Basins

While a state-wide water sharing agreement may be limited by geography, basin-wide water sharing agreements have been highly successful in Colorado. Operators in the Piceance Basin have profited from a water sharing agreement coupled with water recycling facilities that have allowed them to reduce reliance on freshwater resources as well as operating costs. Figure 5 plots water volumes consumed in the Piceance Basin for HF alongside wastewater production volumes per year. Of note, the majority of the HF water volumes may be sourced from recycled waters, but the COGCC does not currently track sources of water. The plotted data shows that the basin produces more than enough wastewater to provide ample volumes of water for HF completions.

Figure 5. Water volumes used for HF and Total Produced Water Volumes, Piceance Basin. The plot presents sums of water consumed for HF operations (‘Frac Water’) and overlays the annual sums of produced (waste)water volumes, as reported to COGCC by operators.

Data for the Denver-Julesburg (DJ) Basin is presented in Figure 6, below. The plot shows that while the HF operations in the basin have in certain years consumed drastically elevated volumes of water, wastewater volumes would still be able to offset the majority of HF water requirements. Hydraulic Fracturing operations in the DJ Basin consume the majority of frac water in the  state, and the drastic increase in consumption 2017-2020 and then 2021 is particularly pronounced. Similar trends are viewable in the county-specific plots, presented in Appendix A. The plots show that the majority of HF water consumption in the DJ Basin is driven by Weld County (Fig 1A), and the majority in the Piceance Basin is driven by Garfield County (Fig 2A).

Of note, the sharp decrease of HF water consumption volumes in 2020 can be attributed to the market downturn during the Covid-19 pandemic. Regardless of Covid-19, the HF water consumption volumes have trended down since 2018 for both basins and most counties. This trend is likely an economic correction considering the lack of substantial response in oil production to extremely high HF injection volumes. Adams County HF volumes are an outlier of these trends (see Fig 5A), as extraction development encroaches on suburban communities surrounding Denver. Suburban and even more urban drilling will likely increase as oil companies (operators) explore for reservoirs farther from the established and nearly exhausted sweet spots of the DJ Basin.

Figure 6. Water volumes used for HF and Total Produced Water Volumes, Denver-Julesburg Basin. The plot presents sums of water consumed for HF operations (‘Frac Water’) and overlays the annual sums of produced (waste)water volumes, as reported to COGCC by operators.

Wastewater Production

The following plot analyses limit the full state datasets just to wells that have undergone HF completions, according to the FracFocus database. The plots of wastewater production volumes are limited to just the three-year period following the HF completion of each well, presented in the plots in one-year increments. To focus on just HF wells the FracFocus database was merged to COGCC monthly production data, then the monthly production data was filtered to include the subset of operational production wells that had completed HF operations. This particular method was employed to assess the contribution of flowback and produced water volumes specifically from HF wells. While FracFocus reports a total of 13,651 hydraulic fracturing events for the 2012-2021 timeframe, a total (N=11,831) hydraulically fractured wells had reported production data, according to the COGCC, and were included in this analysis. As above, plots were created for the full state data and for individual basins. Additionally, plots for individual counties are included in Appendix A.

Annual wastewater production volumes from HF wells for the full state are shown below in Figure 7. While the trendlines for wastewater volumes from just the HF wells show a more visibly correlated relationship to HF water consumption than wastewater volumes for all production wells, the increased volumes of HF water consumption 2019-2022 are barely reflected in the wastewater production volumes. At the elevated water consumption volumes there is a clear disconnect between increased water consumption and increased oil production.

Figure 7. HF Water Use and Wastewater Production from HF Wells, Statewide. The plot shows the volumes of wastewater produced from wells during the first, second, and third years following HF completion, overlaid with water consumption (‘Frac Water’) of those same wells.

Trends in the statewide data show that when injection volumes are unreasonably high, such as during the year 2018, the volumes of wastewater produced comprise a smaller percentage of the volumes injected. Beyond 2019 only the ‘Year 1’ estimates should be considered accurate, due to incomplete data, and the produced water volumes for 2020 and 2021 are likely under-reported. Table 2 below presents the data points used to generate the graph, and reports the percentage of “returned” produced water volumes to HF water consumption volumes. As seen in the table below, 2019 is a distinct outlier. If the incomplete data (2019-2021) is dismissed, the data shows that wells that undergo HF completions in Colorado return an average 30% of injected water volumes in the first 12 months following HF completion, 38% after 24 months, and 44% after 36 months. While these volumes may not fulfill the consumptive needs of all HF completions, these production wells represent just a fraction of the produced water volumes generated in the state.

Table 2. Hydraulic Fracturing Water Injection Volumes and Returning Produced Waters. Volumes of water used for HF completions and volumes of produced waters returned from those wells are presented in the table below. Volumes of produced waters returning 1, 2, and 3 years following HF completion are shown, along with the percentage of the original HF volumes each year represents.

Individual Basins

Separating the DJ and Piceance Basin HF water consumption and wastewater production data shows similar trends to the separation of total produced water volumes for each basin, discussed above. The DJ Basin, shown in Figure 8, is largely responsible for the gap between HF water consumption volumes and wastewater production volumes. In the Piceance Basin the majority of water consumed by HF operations returns to the surface as wastewater, even following the disproportionate volumes used for HF from 2017 through 2019, more than half the injected volumes were reported to have returned to the surface as produced water volumes. Figure 9 below visually demonstrates that relationship in the Piceance Basin.

Figure 8. HF Water Use and Wastewater Production from HF Wells, Denver-Julesburg Basin. The plot shows the volumes of wastewater produced from wells during the first second and third years following HF completion, overlaid with water consumption (‘Frac Water’) of those same wells.

Figure 9. HF Water Use and Wastewater Production from HF Wells, Piceance Basin. The plot shows the volumes of wastewater produced from wells during the first, second, and third years following HF completion overlaid with water consumption (‘Frac Water’) of those same wells.

Impact of Increased HF Water Volumes on Oil Production

While injection volumes of water consumed during HF completions varied widely by basin, there is a conserved trend showing that the response of oil production to extreme volumes of injected HF water is largely inconsistent. Figures 10 and 11 present the production and HF water use data for the entire state of Colorado and just for the DJ Basin. The similarity in the plots makes it clear that nearly the entirety of oil production volumes reported to the state are produced from the DJ Basin.

In both plots below (Figures 10 and 11), oil production from 2013 through 2016 aligned closely with the consumed HF water volumes.  In 2017, volumes of water used for HF completions rose dramatically, while oil production lagged far behind, and this trend disconnecting HF water use from oil production continued through 2022. A similar trend was visible in the data for the Piceance Basin, where spikes in the consumption of water for HF activities had little impact on oil production volumes. In the case of the Piceance both volumes of water consumption and oil production were nearly an order of magnitude smaller. The data shows that the extreme volumes of water that were consumed for HF operations since 2017 did not generate a corresponding increase in oil production of comparable magnitude. Of note, many of the HF operations in producing formations on the Western Slope target tight gas rather than tight oil.

Figure 10. Oil Production and HF Water Consumption, Statewide. The plot shows annual volumes of water consumption for HF completions (‘Frac Water’) and oil production volumes from just wells consuming water for HF completions.

Figure 11. Oil Production and HF Water Consumption, Denver-Julesburg Basin. The plot shows annual volumes of water consumption for HF completions (‘Frac Water’) and oil production volumes from just wells consuming water for HF completions.

Figure 12. Oil Production and HF Water Consumption, Piceance Basin. The plot shows annual volumes of water consumption for HF completions (‘Frac Water’) and oil production volumes from just wells consuming water for HF completions. Note that much of the HF activity on the western slope targets tight gas rather than tight oil.

Existing Sources of Pollution

In addition to the threat of water scarcity, the development of produced water recycling infrastructure can be completed in a manner that reduces the uncontrolled releases of pollutants that cause local and regional air quality degradation. Currently the vast majority of produced waters are re-injected underground using class II disposal wells, and recent inspections by certified thermographers show that class II injection disposal facilities are considerable sources of uncontrolled emissions.

Collaborating with FracTracker on the Front Range, Earthworks inspected class II injection well facilities using a FLIR GF320 optical gas imaging camera. Uncontrolled emissions were documented at six of the twelve sites visited. Footage from five sites are presented in the story map below. The emissions include toxic and carcinogenic volatile organic compounds (VOCs) that degrade local air quality and are the dominant precursor to ozone on the Front Range, in addition to potent greenhouse gasses such as methane. A formal complaint was submitted for each site, but none of the documented leaks were considered violations by the Colorado Air Pollution Control Division. The calculated VOC emissions from the storage tanks where we observed uncontrolled emissions were reportedly below the threshold for control measures or regulatory oversight, nevermind there were multiple tanks on each site. This is a major loophole for oil and gas operators in Colorado and is a major contributing source to ground-level ozone creation on the Front Range. Each of these documented emissions events would be blatant violations in states like California that are focused on ozone reduction, greenhouse gas reduction, and protecting communities from leaking carcinogens.