How Recursive Predicts and Prevents Groundwater Pollution

Why Groundwater Protection Matters

Groundwater is a vital resource that supports communities, agriculture, and industry by providing clean water for everyday use. But when it becomes contaminated, the consequences can be serious, ranging from health risks due to chemical exposure to ecological damage and costly clean-up efforts. Ensuring the safety of groundwater for human consumption and managing it responsibly is critical for protecting public health, sustaining ecosystems, and securing reliable access to clean water.

When planning a new industrial site, it is essential to understand potential risks to the water safety of surrounding communities. Similarly, in cases of existing industrial leaks, it's vital to project the situation one, five, or ten years into the future to develop effective mitigation plans if communities are negatively affected.

However, high-resolution measurements of dissolved substances in groundwater (e.g., nitrates, PFAS, etc.) can be very expensive. The cost of drilling a well to measure concentrations at relevant depths (typically tens of meters underground) can reach $5,000 according to our market research (1, 2, 3). Therefore, a technology capable of accurately recreating the complete pollution profile from a limited number of measurements is highly desirable.

That’s where Recursive comes in. We’ve developed an AI-powered groundwater pollution model that uses just a few concentration measurements to identify the pollution source and extrapolate contamination levels across the entire area of study.

How the Model Works

Our process begins with readily available input data: groundwater pressure data from open repositories (e.g. gbank Water Environment Map, gbank WellWeb GSJ), along with key hydrological properties of the region, such as hydraulic conductivity (Figure 1). These can be sourced from public datasets or provided directly by the client.

Figure 1. Conceptual system diagram of the solution. Hydrological data and client’s measurements are input into a trained AI model, which identifies the source of pollution and predicts future leak dynamics, enabling and informing business decisions related to pollution management.

These properties determine the dynamics of the AI model, allowing for accurate prediction of how pollutants would be transported through the groundwater.

Once the general dynamics of the area are established, we adapt the model to the specific case using the client’s field data - specifically, measurements of pollutant concentrations taken within the contaminated area.

To illustrate this, we show a simulated example of pollutant spread in a flat region, with a realistic pressure gradient and hydraulic conductivity typical of a gravel-sand mix (Figure 2). This could represent an industrial leak on flatlands near a mountainous area.

Figure 2. Spread of solute in groundwater with time, simulated by a trained AI model. Red sphere: source of the solute (pollutant), surfaces show constant concentrations (magenta: 1.0 mg/L, green: 0.5 mg/L, blue: 0.1 mg/L, red: 0.02 mg/L). Arrows show the direction and magnitude of the volumetric flux q=-K∇p, m/day.

In this mock example, we simulate a customer taking measurements at arbitrary locations, then use those measurements to infer the source of contamination. The setup reflects a realistic scenario of a pollutant spill in a flat terrain.

Measurement locations can be chosen at random. We estimate that a localization accuracy of 40 meters can be achieved with 10 or fewer measurement points.

The model ingests the sparse measurements of concentrations and gives the most likely parameters of the pollutant leak, including the most likely location of where the leak occurs (Figure 3).

The model is trained to take into account a diverse set of hydrological properties and works well with various groundwater profiles, streamflows, and heterogeneous hydrological properties of the ground.

Figure 3. Measurement and location inference process. 1. Wells are bored (black) and measurements are taken at arbitrary locations (green circles). 2. The AI model ingests the concentration data from the measurement points, and identifies regions where the source location is probable. Here we show the region of low probability (pink), low-intermediate probability (green), high-intermediate probability (blue), and high probability (red). 3. A priori unknown true location of the source (red wireframe) is close to the high-probability region identified by the model.

Finally, once the source is identified, the model reconstructs the full pollution profile for the area of interest and predicts how the pollutant will spread over time. These insights help guide the client’s next steps, whether that’s containment, extraction, or other actions to manage and reduce the contamination.

Potential Use Cases: Who This Technology Helps

Our modeling approach can be applied in a variety of contexts, involving different solute substances and industries. In the sections below, we organize the potential applications in three ways: first, by industry; second, by business objective; and third, by relevance to water management and regulatory compliance.

Applications by Industry

Here are several specific industry contexts, associated contaminants, and possible business applications where our model can be effectively used:

Agricultural Runoff:

• Model the transport of nitrates, pesticides, and other agricultural chemicals from fields to groundwater.
• Create effective management strategies to reduce agricultural runoff, such as model the effect of underground walls on runoff containment.

Mining Impacts:

• Model the release and transport of heavy metals, cyanide, and other contaminants from mining operations.
• Design and evaluate strategies for mine water treatment and disposal.

Industrial Discharges:

• Model the fate and transport of solvents, petroleum hydrocarbons, and other industrial chemicals in groundwater.
• Evaluate the effectiveness of containment measures and treatment technologies for industrial discharges.

Emerging Contaminants:

• Model the transport and transformation of emerging contaminants such as pharmaceuticals, microplastics, and per- and polyfluoroalkyl substances (PFAS) in groundwater.

Radioactive Waste Disposal:

• Model the long-term migration of radionuclides from underground radioactive waste disposal facilities.
• Assess the potential risk to human health and the environment from radioactive contamination of groundwater.

Nutrient Transport:

• Simulate nutrient transport to surface water (e.g. phosphorus, nitrogen), to model eutrophication, e.g. excessive algal growth and oxygen depletion in water bodies.
• Assess mitigation strategies, as well as future risks.

Landfill Management:

• Localize landfill liner breakages.
• Model leachate migration from landfills to ensure liner integrity and prevent groundwater contamination.

Carbon Sequestration:

• Model the movement of injected CO2 in underground reservoirs for carbon sequestration projects.

Geothermal Energy (with additional engineering):

• Model heat transport in groundwater systems for geothermal energy production.

Applications by Business Objective

Here are several use cases focused on environmental impact reduction and contamination management. These activities are broadly applicable across the industry-specific examples above:

Remediation Design & Optimization:

• Design and optimize pump-and-treat systems, permeable reactive barriers, monitored natural attenuation strategies, or other remediation techniques.
• Predict the effectiveness of different remediation scenarios and select the most cost-effective approach.
• Optimize well placement and pumping rates for plume containment or extraction.
• Find optimal positioning of underground walls for pollutant containment.

Long-Term Monitoring:

• Optimize the design of groundwater monitoring networks (wells, IoT, remote sensing) to effectively track contaminant plume migration and remediation progress.
• Predict future contaminant concentrations to assess the long-term effectiveness of remediation efforts.

Brownfield Redevelopment:

• (definition): Brownfield is previously-developed land that has been abandoned or underused, and which may carry pollution, or a risk of pollution, from industrial use.
• Assess the extent of contamination on brownfield sites to determine remediation requirements for redevelopment.

Source Identification:

• Trace the origin of a contaminant plume to identify the responsible party or source area.

Applications for Water Management and Compliance

These applications focus on managing water resources responsibly and supporting regulatory compliance:

Wellhead Protection:

• Delineate wellhead protection areas for public water supply wells to minimize the risk of contamination.

Aquifer Storage and Recovery (ASR):

• Model the movement and mixing of injected water in ASR projects to optimize storage and recovery efficiency.
• Evaluate the potential for geochemical reactions between injected water and native groundwater.

Permitting:

• Support applications for permits related to wastewater discharge, underground injection, or other activities that may impact groundwater quality.
• Demonstrate compliance with regulatory requirements for groundwater protection.

Regulatory Decision Making:

• Inform regulatory decisions related to groundwater management and protection.
• Evaluate the impact of proposed regulations on groundwater quality.