Groundwater Models
Groundwater Withdrawal Feasibility Analysis to Identify Sustainability Scenarios in the Sotaquirá River Basin, Colombia
Researcher: Luis Felipe Sierra Ponguta
Mentor: Dr. Andreja Jonoski
Supervisor: Dr. Gerald A. Corzo
Abstract
Climate change is a phenomenon that has generated changes in the water cycle at a global level. In Colombia, climatic variability is increasingly impacting water resources. In the case of the Chicamocha river basin, which is located in Boyacá state in the Andean region, this phenomenon decreases the water flows by 30% (García et al.,2010). Reduction of water flows in the Chicamocha river basin and the increasing water demands for different productivity sector usages may compromise the availability of water in the future. The Sotaquirá sub-basin, located in the upper Chicamocha river basin has a bimodal flow regime where the first trimester is the most critical, with minimum discharges under 2 m3/s.
Water from the upper Chicamocha river basin used for different economic purposes. The system has a high stress due to the Usochicamocha irrigation system's presence and the need for water supply for different cities. Gensa and Termopaipa are two coal-based thermal energy generating plants located on the Chicamocha river that provide energy to other cities within the river basin. Coal combustion produces energy by moving turbines using water vapor pressure. The upper Chicamocha basin has high stress due to high water consumption for different economic sectors. Recently, the population's energy demands have been augmented significantly, so increasing the intake flow to upturn energy production is necessary. Current total water demands in Gensa and Termopaipa are 0.5 m3/s; however, future demands will be increased by 0.3 m3/s, which means high stress on surface water in dry periods. Vulnerability Water Index was computed to identify the water system's degree of fragility to maintain the water supply in periods of low water or events as the El Niño phenomenon. The majority of sub-basins that make up the Chicamocha river basin have high vulnerability in drought conditions. Surface water is used for different purposes, and low flows become critical even with a reservoir in the basin. High stress in the current and future water demands makes it necessary to study groundwater availability for critical periods such as shortage and droughts.
The purpose of this work is to analyze, assess the water availability, and propose a sustainable alternative to surface water intake, by using withdrawal of groundwater from the Sotaquirá river basin to supply thermal energy production in drought periods. This zone is part of the Chicamocha river basin, and the performance analysis is focused on a sub-basin with high groundwater productivity. To evaluate groundwater availability in the region of interest, the Sotaquirá river basin was selected due to the geographic location and good water yields (baseflow) in droughts. This basin, whose area is 140 km2, supplies surface water for the Usochicamocha irrigation district and other uses, including agriculture, urban water supply, and energy generation. The groundwater analysis will be performed at a regional scale to evaluate the possibility of groundwater extraction from the Sotaquirá river basin, located near the hydrothermal power plants.
The importance of this study lies in proposing alternatives for water extraction from the aquifer and studying the effects on the base flow downstream of the basin. To propose alternatives, it is necessary to develop suitable groundwater models, and to understand the most sensitive variables in the development of regional groundwater models. By understanding the groundwater hydrodynamics, we can accurately estimate the effect of withdrawal in the aquifer's water. The geology of the Sotaquirá river basin is complex. However, groundwater withdrawal has a high potential due to primary and secondary porosity aquifers (POMCA, 2006). Some parts of the basin have limited recharge due to the presence of impervious rock formations that reach the surface and cover some areas of the basin. On the other hand, the geology's stratigraphic complexity in the zone makes some aquifer lateral discontinuities, and the thickness may vary along the basin.
Problem description
The upper Chicamocha River Basin (CRB) is a water system considered as a significant water source for supplying water demands to different uses, including agriculture, human
consumption, and electricity generation. Currently, the water system can sustain the current demands for the different activities through mainly exploitation of surface water.
Within the basin are allocated Termopaipa and Gensa coal-fired power plants, which use the water extracted from the river to produce energy through the vapor pressure produced by heating water. The current combined consumption of these plants is 550 l/s. However, current trends show that for a horizon of 20 years, 850 l/s will be necessary to supply the region's energy needs. The over-stress that the system currently has, mainly in times of drought, makes it necessary to evaluate sustainable alternatives for groundwater use, close to the study area, allowing for the future increase in demand for energy needs. The historical hydrological regimes show that the first four months of the year (January - April) cover the period of low flows in the basin. However, the historical minimum flows show critical conditions in all months of dry years.
The Sotaquirá River Basin (SRB) is a tributary to the Chicamocha River and was selected for the analysis due to its location upstream of the power generating plants. The lower part of the basin has hydrogeological characteristics indicating potential for groundwater development, especially due to predomination of high conductivity alluvial soils. In this research a groundwater model was built for the entire basin (parent model) where the general groundwater drainage condition has been analyzed. This was followed by development of a specific model for the study area (child model), in which the withdrawal alternatives are analyzed.
Two alternatives were analyzed at time horizons of 10 and 20 years with demands of 150 l/s and 300 l/s, distributed in 6 wells located within the zone with the highest groundwater yield under two withdrawal options. These scenarios have four-month length (dry period) for each year and sixteen-month length (dry year plus annual dry period) and were evaluated for the period (1990 - 1993), which matches to historically driest years recorded in Colombia and the study basin.
The alternatives' analysis shows that for withdrawals of 150 l/s (Alternative 1) for options 1 and 2, the groundwater head drawdown and the effect on the reduction of base flow in the river are less severe, with a groundwater drawdown near to 9 meters, base flow reduction of about 24% are expected, and the remaining base flow is 428 l/s. Alternative 2, in which the pumping rate is 300 l/s, for the option 2 causes a groundwater drawdown of about 16 meters, a reduction of the base flow of about 45% and remaining base flow of 222 l/s. As environmental flow for the SRB is 390 l/s, Alternative 1 was selected as the most sustainable for the system since recharge favours groundwater recovery
Modelling solution
The analysis to be carried out requires integrating the different components that are part of the groundwater balance. Since main input to the groundwater system is natural recharge, which is the result of the mass balance of the physical processes that occur in the surface and sub-surface, one key methodological step associated with the groundwater model development is recharge estimation. To achieve the posed research objectives, the following main methodological steps are proposed: Low flows analysis in CRB and SRB: Low flows are a susceptible component within the water system. The occurrence of low flows is associated with periods when the basin has higher water stress. One of the effects of groundwater abstraction near streams is the infiltration of the flow into the aquifer, reducing the base flow. The estimation of the environmental flow is vital to know the impact that the reduction of the base flow product of pumping can have on aquatic ecosystems and other economic activities that depend on surface water. Analysis period selection: The historical information from records of climatological stations and flow meters allows the analysis of critical periods in which droughts have occurred in the area. The flow records are helpful to know the seasonality of the low flows and the average duration that these have in the basin, which allows knowing the periods in which the abstraction of the auxiliary system will be necessary. Phenomena related to El Niño, La Niña, and El Niño-Southern Oscillation (ENSO) have produced historical indices of climate change that allow us to know the historical years in which global temperature changes have occurred and therefore in the study basin. The analysis will be carried out in a year where the Niño phenomenon has historically occurred. Recharge Computation: This parameter requires the estimation of the water balance in the area of interest and the interaction of the hydrologic physical process in the basin. A comparison of different methodologies will be made for its estimation, and one will be selected to provide values of this input parameter to the groundwater model. Base flow separation: The base flow corresponds to the hydrograph's part coming directly from the aquifer to the surface. The groundwater model simulates the exchange between the channels and the aquifer through the base flow, which is very useful for calculating the river packages within the model. From the calculation of the effective recharge, which requires the calculation of the base flow daily for a long time, the base flows for the analysis period are estimated.24 Hydrogeological Analysis: The groundwater drainage conditions depend directly on the lithological structures that underlie the soil. The geological complexity that the basin may have makes it necessary to create a conceptual geological model using the information from geological maps available in the area. From this model, the discretization is carried out in a mesh whose cells can be associated with the geological structures' respective conductivities. Groundwater model development: From the mesh built from the geological model, importing the groundwater model to assign initial and boundary conditions is possible. This model will represent the current equilibrium system in current conditions (undisturbed). Two groundwater models will be needed for the analysis. Parent model will drop a general overview of the basin’s groundwater drainage. Child model will be necessary to perform the groundwater abstraction analysis by reducing uncertainty in the boundary conditions and input variables, and improving computational times. Geospatial and field data will be used to develop the model and to assign required boundary and initial conditions. Sensitivity analysis and model calibration: Sensitivity analysis will be performed in the Child model to estimate the input variables to be calibrated in the model. Available measurement records will be used for model calibration. The calibration will be carried out in a transient state, and the parameter adjustment will be carried out manually. Simulation of the effect of withdrawal from the system: From the calibrated Child model, a set of hypotheses for well location and pumping rates will be assessed to determine the most optimal solution with less adverse effects. Recovery system analysis: Analysis will be performed of the recovery capacity of the system by natural recharge during wet conditions, and, potential future measures to maximize the storage of water in the system. Land subsidence analysis: The effects that water extraction on the physical composition of the soil may be significant. An initial estimation of land subsidence rates due to the effect of withdrawals in the groundwater layers will be performed for the case study
Conclusions
The analysis carried out in the Sotaquirá river basin corresponds to an initial representation of the groundwater flow conditions, which is still quite approximate, due to the limitations of the available information. The absence of piezometers with historical level records to perform the model's calibration makes it necessary to carry out more exploratory analyses in the area to determine the hydraulic parameters of the subsoil more closely. The changes in the aquifer levels caused by the effects of recharge and discharge have a direct influence on the base flow of the channels that drain through the system. The reduction of the groundwater level caused by the continuous abstraction by wells close to channels causes the channel water's infiltration into the aquifer, reducing the base flow. The proposed scenarios directly affect the base flow, causing minor drainage in the rivers that pass through the abstraction area. The effects of continuously withdrawing water during a dry year directly affect
economic activities and ecosystems that benefit from surface water.
From historical flows recorded in the basin and antecedent analysis, it was concluded that one of the most critical historical periods recorded was 1990 – 1993, which had El Niño phenomenon conditions. From a seasonal analysis of low flows and from daily records the driest period of the year is January – April.
The environmental flow in the SRB was calculated from the flow duration curve of the Maguncia station. The analysis drops a value of 33696 (390 l/s).
An estimation of the recharge was carried out through different methodologies, being the hydrological analysis daily in the HEC-HMS model, using the Soil Moisture
Accounting loss method, the most appropriate due to the possibility of calibrating the model from of flow records observed on the Sotaquirá River. The results show that in most years there are no recharge in the first trimester.
The groundwater child model's sensitivity analysis shows that the most sensitive parameter is recharge; however, this parameter was calibrated from the runoff rainfall
model, reducing its uncertainty. Hydraulic conductivity, mainly in layers 1 and 2, assigned as convertible, has a high sensitivity since groundwater levels oscillate within
this layer’s elements.
The calibration was performed for the base flows using the Stream Flow Routing (SFR) package, which allows the transit of measured flows in the river sections that interact with the aquifer. The uncertainty of the underground parameters influences the SFR package's calibration, given the interaction between the systems. The calibration showed a better performance of the model for low base flows.