How Grounded AI Could Save $1 Billion in the Permian Water Sector
The oil and gas industry’s recent and explosive success in the Permian Basin has been a boon to production and profitability for producers as well as to American energy independence overall. Production in western Texas and southern New Mexico is expected to double its current production rate to about 8 MMbbl/d in four years. For perspective, that estimate surpasses the U.S.’s total production numbers from six years ago. But with all of that opportunity, a growing water challenge must be taken into account when operating and investing in the Permian Basin.
At today’s oil production rates of more than 4 MMbbl/d, assuming an average water-oil ratio of 5:1 inside a range 1:1 in the Midland Basin and 10:1 in the Delaware Basin, this record-breaking area of drilling activity is producing 20 MMbbl/d of produced and flowback water.
It is estimated that 10,000 wells will be hydraulically fractured in 2019 in the Permian Basin. With laterals exceeding 10,000 ft in some cases, the average fracturing operation requires 500,000 bbl of water per well. This results in water usage for fracturing purposes at more than 5 Bbbl/year.
Current estimates indicate that 13% of the water used for fracturing purposes falls in the reuse category and runs through various treatment technology solutions. The bulk of the remaining water supply is sourced from groundwater wells interspersed across an active landscape of producing wells.
Opportunity in water
Across the industry, from a life-of-well perspective, water has long held the less than enviable pole position on line item costs. In the Permian, water is extending its lead on the cost side of the equation due to volume and, for that reason, is in the spotlight for producers, regulators and the growing sector of water infrastructure solution providers. Some question the sustainability of this commodity production base. Others see an incredible opportunity and anticipate that technological innovation will rise to the challenge. Based on the industry’s track record of reinventing itself, this too will be solved.
If the Permian is to reach the 8-MMbbl/d production forecast, the industry needs to do more than access additional capital to expand the asset base. Part of the solution resides in an underutilized asset that touches every process in the water management cycle.
Drilling activity in the Permian has attracted more investment in water infrastructure than any other basin in the country and perhaps the world. This large-scale investment, primarily in pipeline transport, disposal and treatment, has reduced the cost-per-barrel unit cost of water management. Incremental investment in pipe and treatment will continue to put downward pressure on costs. However, there are other counter forces at play. Anticipated ever improving economies of scale for disposal may be approaching a saturation point. Without new sources of performance improvement, these offsetting trends, combined with steep growth in volume, will drive water costs higher. The primary cost drivers are disposal, treatment and transportation.
Disposal: Saltwater disposal wells—the disposal option of choice in the Permian—perform better in this basin from a barrels-per-day perspective than in most other markets. If the disposal wells cannot be utilized in an efficient manner that keeps pace with production, well productivity can be slowed or halted assuming there are no alternative transport/disposal options available. This puts the pressure on the injection well facility operator to perform, especially in high-volume and disposal-capacity-constrained markets.
Treatment: Reuse efforts in the Permian are on the rise in response to the growing demand for frac water and permitting headwinds on injection wells. Treatment costs can be higher than in other basins due in part to the naturally high salinity rates in the groundwater. Total dissolved solids (TDS) levels often exceed 100,000 milligrams (mg), limiting the spent water’s industrial or agricultural beneficial reuse options. In other basins, such as the Powder River Basin in Wyoming, produced water TDS is as low as 15,000 mg. The other complicating factor for treatment is water composition variability driven by elements other than salt. This complicates the quest for scale by the treatment provider and places a premium on collaboration between producer and treatment partner on water composition data. More accurate and timely data can be used to improve system productivity.
Transportation: The density of drilling operations in the Permian has enabled significant investment in pipeline transport. This growing network of pipe combined with the migration of disparate producer-owned systems into broader midstream ownership models has helped provide cost certainty at reduced risk to the producer. Truck labor shortages, as well as increasing road safety concerns, serve as qualitative incentives for additional pipe investment. As pipeline networks have expanded, cost savings opportunities have surfaced in relation to system optimization. The optimization decision involves water flow by source, transport route alternatives and disposal or treatment points in the network.
What else can be done?
Investors are aware that investments in increased productivity will make the Permian more competitive and deliver more resiliency to cyclical commodity prices. On the production side, there is a surge of new technology solutions, including software that centers on subsurface data analytics as a source of value.
As these improvements drive well productivity, the spotlight on the need for offsetting improvements in water management will persist and grow. Industry analysts have speculated that as much as $9 billion will be needed over the next decade to dispose of produced water. While private-equity investors have shown a willingness to fund appropriately contracted water projects with investment capital, there is an opportunity to improve capital efficiency and asset operations across the water management value chain.
The water sector has invested in data collection and facility monitoring systems. The opportunity lies in monetizing these data, perhaps the most underutilized asset in the water sector. Similar to advancements in production, there is an opening to apply analytical techniques to water management strategies and a library of tools that target water facility processes and capital deployment.
What could 5% or even 10% in water management cost savings mean for the world’s most productive oil field?
Simply put, grounded artificial intelligence (AI) doesn’t have its head in the clouds. Grounded AI is hyperfocused on specific, practical, data-intensive tasks that help water asset operators surface valuable and actionable insights about their operations from the data they already collect. Grounded AI solutions in water are designed to encourage water transport, treatment and disposal operators to tie hard return on investments (ROIs) to all AI investments by specific use cases that enable their business needs, harness the value of existing assets and process data, be quick to implement, and easy to understand and use.
Innovation in the oil and gas industry has been remarkable in many ways over recent decades. Inside the data science component of this innovation, there are areas of the business that are far to the right on the maturity curve and other areas that are lagging. The oil and gas industry as a whole has committed significant investment to data collection with a focus on drilling and production. Water represents the next frontier for data science in oil and gas. Water operators have invested in SCADA systems, control centers and other Internet of Things solutions. However, most of the sector has been focused more on growth and keeping up with customer demand than monetizing the data asset. This is especially true in the Permian given the current and forecast rate of growth. The powerful point about this underutilized data is that it touches every process of the business. Surfacing the insights locked inside the various sources of raw data with predictive analytics, automated alerts and/or more robust optimization analysis has the potential to drive bottom-line savings and broader business transformation.
For example, one of the clearest ways for water disposal and treatment operators to see proof of ROI from grounded AI is by preventing equipment failure or a costly equipment-related risk event via predictive modeling. It is crucial that injection well field crews are equipped with operational intelligence to adjust workflows in advance of faults, avoid downtime and maximize the daily permitted capacity of their active wells. These types of proactive solutions also can drive savings by reducing the need for redundant equipment.
Predictive modeling is a top priority for operators that need to be able to take immediate action on collected data. In Plutoshift’s recent research, 76% of midlevel executives across industries said they need software solutions that analyze data in real time to take immediate action on collected data.
Working off of spreadsheets and other reports after the fact exposes the operator to unplanned outages, maintenance costs and, more importantly, customer disruptions. This is a known sub-optimal position for operators that would prefer to know if a problem might occur so they can take the actions necessary to address equipment and asset issues before they reach the point of failure. But few have made this transition to proactive solutions. Of those same survey respondents, only 12% said they are able to take action on their data insights automatically.
Grounded AI solutions can streamline and automate previously manual data collection and analysis processes like Excel spreadsheets and other data management systems. Labor-intensive data collection and analysis can be automated by a system that is capable of managing millions of microdecisions per hour while providing operators with real-time proactive insights into their processes.
Grounded AI solutions can drive savings with
- Improved maintenance strategies: Predictive or condition- based maintenance or condition-based and equipment fault warning alerts can optimize labor costs in line with accepted risk parameters;
- Manual equipment intervention effectiveness and field crew monitoring: By integrating labor tracking data and facility performance, correlations and data patterns can be detected to identify sub-optimal performance, training opportunities and potential safety risks; and
- Data verification and anomaly flagging: Continuous analysis of operational data can spot anomalies that could lead to the detection of faulty sensors.
Improved system optimization
Midstream water management systems that include multiple transport, disposal and treatment assets are expanding across the Permian. These systems present the operator with incremental value in the form of optionality. Data and modeling are the keys to the kingdom of monetizing the value of the network through performance optimization and risk mitigation. System optimization that incorporates the use of grounded AI can drive value on several levels for the midstream system operator interested in a data-intensive, “moneyball” outcome.
By establishing automated real-time alerts across an interconnected system of disposal facilities, operators can direct water flow via a pipe to avoid system hot spots (i.e., disposal facilities where a pump failure has been predicted or where other complications have caused tank volumes to rise above target levels).
A systemwide optimization model that incorporates a full spectrum of data including inflows by well pad, disposal facility pressure data, water composition data, treatment flows and storage capacity could enable disposal and treatment at lowest costs per barrel by redirecting water in line with facility cost structures or water composition data.
From a strategic perspective, operators can combine grounded AI tools connected to facility operations with contextual data (i.e., geological, permitting and fieldwide water production data from B3 Insight) to better utilize existing assets and plan for system growth via new injection wells or treatment and storage. Savings are recognized in more effective system optimization and efficient capital deployment. A custom, user-defined data monetization platform supported by advanced insight detection tools can enhance existing data collection systems by unlocking and surfacing valuable information that is immediately actionable and ROI-based. Operators can use the centralized analytical platform to instantly generate reports on key performance metrics that impact a facility’s or system’s bottom line, such as cost-per-barrel disposal operating expenses, water volume, chemical costs, skim oil levels, water treatment costs, energy costs, uptime, unplanned outages or equipment faults, and safety events.