Aquifer Modeling
Groundwater represents a fundamental part of the hydrological cycle, the exploitation, efficient management and proper understanding of which represents a safe, reliable and cheap source of water. Nowadays, deficient industrial, agricultural and environmental practices have caused a multitude of aquifers and aquitards to be impacted by organic and inorganic compounds through man-made activities, the immediate consequence of which entails a progressive degradation of the groundwater and with it, the environment.
The characterization and management of the underground water resources involves an integral analysis of a set of tools to determine the characteristics of an aquifer (type and geometry; piezometry, quantification of hydraulic and transport parameters, water balance, level of quality and exploitation, etc.). In the last decade, a fundamental part of any hydrogeological study must be sustained by a numerical model for groundwater flow and contaminant transport that has multiple applications but, to name but a few, a model is useful for:
- Estimating the performance of extraction wells
- Analyzing the interference in a field of wells
- Designing systems of drainage in mines and depletion of the water level in excavations
- Determining collection areas
- Outlining aquifer protection perimeters
- Simulating the impact of contaminants in the saturated area and the relationship with the unsaturated area
- Simulating the transitory state of a dissolved-phase plume and its relationship with the natural environment and other contaminants.
- Assessing the natural attenuation that has been monitored in contaminated aquifers
- Assessing aquifer remediation systems
- Modeling the management of underground water resources
What is a model?
A model is defined as the partial or total representation of a natural system. There are different types of models; specifically numerical models are those where the representation of an aquifer (or part of it) is done by means of differential equations (general flow equation, general transport equation and Darcy’s Law), which are solved by (approximate) numerical methods for each element of the domain.
CAM uses a solution method known as finite differences, in which the domain is divided into rectangles (2-D model), or parallelepipeds (3-D model), and the objective consists of solving the differential equations, obtaining the hydraulic head (“h” in meters or masl) and the final concentration (Cf in mg/L) for the center of each cell of the mesh, as a function of time.
To this end, we apply the software Visual MODFLOW, developed by Waterloo Hydrogeologic Inc (WHI), which is a code that is widely accepted by Mexican (CNA, IMTA) and international (USGS, EPA) bodies.
Methodology
In the execution of a numeric model, CAM applies the following framework for analysis:
(1) Statement of the problem
(2) Obtainment of data
(3) Construction of the Conceptual Model
(4) Design of the mesh of the model (discretization)
(5) Assignment of hydraulic and transport parameters
(6) Calibration and Validation
(7) Analysis of Sensitivity
(8) Predictive Simulations
Simulation of the migration of the contaminating plume vs. Time
CAM has developed a variety of flow and transport numeric models for the public and private sectors, that has as its main approach the prediction and simulation of the behavior of dissolved-phase contamination in an aquifer in a transitory state. To this end, the current flow conditions and concentrations of the solute(s) are reproduced and, after calibration, it is feasible to run a simulation model, in order to assess: (A) potential migration routes for the contaminant(s). (B) Transit times for the contaminant going to a receiving water body. (C) Arrival curves “Time vs. Concentration ” for different external receiving water bodies. (D) Design and optimization of the Remediation system. (E) Simulation of the biodegradation (anaerobia or aerobic) and potential natural attenuation of the aquifer.
Advantages
One of the great advantages of the numerical model is that it allows us to assess different prediction scenarios, changing the conditions that act on the system, such as replenishment and extractions. Moreover, we can set out multiple remediation scenarios in order to optimize the restoration system and minimize the number of elements required (injection wells, extraction wells, trenches, etc.), maximizing its output and thus obtaining a high cost -benefit ratio that allows for a smaller investment, apart from giving us very useful elements for decision-making in the management of hydrogeological and geoenvironmental projects.
GI-GO (Garbage In-Garbage Out) Principle
The main limitation lies in the fact that a numeric model requires a great deal of data, meaning that it is necessary to have a wide range of information about the site. Therefore we have to develop a prior conceptual model that is coherent with the reality of the situation, that is supported by data on piezometric levels; concentrations of the solute; determinations of the hydraulic parameters through slug and/or pumping tests; physical parameters and soil mechanics; chemical parameters of the soil; hydrogeochemistry and water balance. It must be emphasized that a numeric model is not a panacea and, if it is not based on an in-depth knowledge of the site, you end up incurring in the error described by the GI-GO
Garbage In- Garbage Outprinciple, in other words: if you put incoherent data in, you will get incoherent results.
LNAPL Modeling
Despite the fact that the MODFLOW software has many applications, it has some limitations, as it cannot simulate every aspect of an aquifer system and its use (of an advanced nature) is circumscribed depending on the scope of each project in particular.
In porous aquifers that have been impacted by spills of hydrocarbons that are lighter than water, it is very common to find that the aquifer is contaminated by a free phase plume, known as LNAPL.
Compounds that are lighter than water and not water soluble, i.e. LNAPL Light Non-Aqueous Phase Liquids, tend to form a layer that lies on the surface of the water, flowing at a slower speed than the real speed of flow. LNAPL are a major challenge for environmental remediation as they represent the greatest risk to a contaminated aquifer.
To assess and manage sites that are contaminated by LNAPL, CAM applies a model developed by the American Petroleum Institute (API) based on the equations of Van Genuchten, Richards and Darcy's Law, known as LNAPL Calculation Tools, with which the characterization of free phases is complemented, determining the true speed of the LNAPL, the volume and degree of hydrocarbon saturation in the soil, final concentrations at distances and specific water bodies that receive the dissolved phase, the change in composition of the source area vs time, as well as the hydrocarbon recovery rate, taking a variety of remediation techniques into account, mainly through trenches, wells with skimmers and Dual Phase Extraction systems.
Application of EPA models
Depending on the scope and complexity of the project type, CAM evaluates the application of a variety of models that do not require the detail of advanced modeling, however they require the simulation of the toxic compound to determine arrival times and expected concentrations for potential external receiving water bodies, in different time intervals. For this CAM applies two models developed by the Environmental Protection Agency (EPA) that have a variety of applications: (1) Optimal Well Locutor (OWL), developed by the National Risk Management Research Laboratory, which is a program that allows us to predect the mobility of the contaminating plume in this transitory and also makes it possible to analyze and optimize the location of a network of monitoring wells to take maximum advantage of the information obtained in each well, on the basis of the flow direction, parameters for the transport and movement of the dissolved-phase plume.
(2) Hydrocarbon Spill Screening Model (HSSM), developed by the Environmental Research Laboratory, that simulates the hydrocarbons (HC) spill and by inputting data into modules (hydrology, hydrogeology, non-saturated zone, characteristics of the hydrocarbon) gives a series of useful graphs that quantify the volume, deep soil saturation profile, HC mass flows, HC arrival curves vs time, for up to 6 receiving water bodies in real distance.
