Groundwater banking: opportunities and management challenges

Transcription

1 Water Policy 16 (2014) Groundwater banking: opportunities and management challenges Robert G. Maliva Schlumberger Water Services, 1567 Hayley Lane, Suite 202, Fort Myers, FL Abstract Groundwater banking is the use of aquifers to store water to balance seasonal or longer-term variations in supply and demand. The large storage capacity provided by aquifers can be a valuable tool for conjunctive use of surface water and groundwater as well as other elements of integrated water resources management. Successful groundwater banking requires favorable hydrogeological conditions to efficiently recharge, store, and abstract large volumes of water. Additionally, groundwater banking is also highly dependent upon water management and operational policies to ensure that stored water is not abstracted by other users and that the water accounting system of the bank remains in balance. Accumulated credits to withdraw water should not exceed the capacity of an aquifer to safely produce the water at the design rate-of-return for the bank. System participants need to have confidence that credits issued for recharge can be safely recovered when needed. Groundwater banking systems can cause significant local adverse impacts to other aquifer users and sensitive environments during recovery periods. Groundwater modeling is required to develop a sustainable management system that accounts for temporal and spatial variations in the impacts of both recharge and abstraction activities. Keywords: Aquifer recharge; Groundwater banking; Integrated water resources management 1. Introduction Aquifer overdraft, whereby groundwater pumping rates exceed the aquifer recharge rate, is a growing threat to water resources and the sustainability of irrigated agriculture and potable water supplies in many areas of the world. Overdraft can be reduced or ended by increasing capture, reducing demands (water use), substitution by surface water or reclaimed water, conjunctive use of surface water and groundwater, and managed aquifer recharge (Harou & Lund, 2008). Conjunctive use of groundwater and surface water affords opportunities to optimize the use of both resources and can be an essential element of integrated water resources management (IWRM). The basic strategy is that surface water is used as the primary water supply when available, and groundwater is reserved for periods when surface water supplies are inadequate to meet demands. With respect to irrigated agriculture, groundwater doi: /wp IWA Publishing 2014

2 R. G. Maliva / Water Policy 16 (2014) pumping capacity reduces risks associated with reliance on surface water supplies and maximizes expected income (Bredehoeft & Young, 1983). Groundwater banking is a type of managed aquifer recharge (Dillon, 2005) in which an aquifer is used to store water to balance seasonal or longer-term variations in supply and demand. The great storage volume available in aquifers is taken advantage of to store excess water for later use (typically surface water), which might otherwise not be put to beneficial use. Groundwater banking also includes regulatory storage-type aquifer recovery and storage (ASR) systems, in which injection of water into any aquifer confers a later right to withdraw the water (Maliva & Missimer, 2008, 2010). Groundwater banking systems typically store water in overdrawn freshwater aquifers. Implementation of groundwater banking has been concentrated to date in the semiarid and arid south-western United States (particularly parts of the states of California, Arizona, and Nevada), where there is a great need for water storage and large available alluvial aquifers. Groundwater banking, as the term is used herein, does not include ASR systems that locally store freshwater in brackish or saline aquifers through a displacement process. The principal value of groundwater banking is that the storage of water in aquifers can help ensure regularity and continuity of water supplies and reduce both water scarcity and the costs associated with scarcity (Greydanus, 1978; Thomas, 1978; Pulido-Velazquez et al., 2004). Economic-engineering optimization modeling performed for southern California demonstrates that groundwater banking provides flexibility for better temporal regulation of flows and can facilitate water transfers that are needed to take economic advantage of conjunctive use (Pulido-Velazquez et al., 2004). Groundwater storage banks, which involve the physical storage of water, are not the same as water banks or water transfer banks, which establish a procedure or process to facilitate the transfer of water between willing sellers and buyers (Miller, 2000). The overall goal of water banking is to facilitate the transfer of water from low-value to high-value uses by bringing buyers and sellers together (Frederick, 1995; Clifford et al., 2004). Water banks have assumed the roles of broker, clearing house, and market maker (Clifford et al., 2004). Brokers connect or solicit buyers and sellers to create sales. Clearing houses serve mainly as a repository for bid and offer information. Market makers attempt to ensure that there are equal buyers and sellers in the market, and they act to create and increase liquidity by ensuring trades occur even when counter parties are not available in the market. Groundwater banking requires an accounting system to track the recharge and abstractions of stored water and may also include a market system to encourage the storage of water and make stored water available to users with the greatest needs (i.e., increase economic efficiency). Depositors would receive or earn credits for the recharge of a given volume of water, which could later be cashed in, for water recovered from the groundwater bank. Banking systems could be designed that allow and facilitate the trading of credits. The water resources benefits of groundwater banking are compelling in terms of capturing and storing flows that may not otherwise be put to beneficial use and optimization of the overall management of water resources. However, the long-term viability of groundwater banking systems requires that the physical aspects of the system (i.e., water storage) and accounting system be congruent. For a groundwater bank to be successful, the bank s depositors must have assurance that water deposited in the bank will be available to them for withdrawal at a later time when needed (Sandoval-Solis et al., 2011). The success of systems will ultimately depend upon the existence of regulatory frameworks and system-specific policies that ensure that credits for recharged water have clear, known, and protected values in terms of the ability of, and conditions placed upon, credit holders to later withdraw water. The rights of credit holders also need to be reconciled with the rights of existing and future groundwater

3 146 R. G. Maliva / Water Policy 16 (2014) users, along with environmental concerns. Groundwater banking systems thus present greater water management and policy challenges than are usually first appreciated. The banking analogy breaks down when one considers that deposits and withdrawals of money at different locations within a monetary bank have no adverse effect on the bank. This is not true for water banking systems, where excessive withdrawals at a location unsupported by the recharge of real water credits can cause significant adverse effects to other bank participants. Another significant issue is that over the long term, water credits can exceed the sustainable yield of the water bank. The objective of this paper is to address some key technical and regulatory issues that affect groundwater banking systems using the results of a theoretical groundwater modeling investigation of the impacts of system operations on water availability. 2. Groundwater storage concepts The underlying concept of groundwater banking is that an aquifer is essentially used as an underground storage tank. Recharge of a given volume of water results in a corresponding increase in the volume of water stored in an aquifer, which would be manifested by an increase in water levels (or heads). In the tank analogy, withdrawal of the same volume of water as recharged would result in a decrease in water levels back to pre-recharge (baseline) levels. However, the tank analogy is a gross simplification as aquifer water levels depend upon a number of other factors including groundwater pumping by other users, leakage, and natural recharge and discharge. The essential feature of groundwater banking is that the recharge of water by either land application or injection wells increases the volume of water stored in an aquifer and that at least some of the water will be recoverable at a future date. Groundwater banking thus needs to be considered in the context of the water budget of an aquifer or groundwater basin, in which inputs of water minus outputs is equal to the change in storage. More specifically, R MAR þ R N þ L ¼ D þ Q þ DS V where, R MAR ¼ managed aquifer recharge, R N ¼ natural recharge, L ¼ net leakage into aquifer, D ¼ natural aquifer discharge, Q ¼ pumping abstractions, and ΔS ¼ change in storage (all units are volume). The change in the volume of water in storage is equal to the product of the change in head (Δh), aquifer area (A), and aquifer storativity (S), DS V ¼ DhAS integrated over the entire area of an aquifer. A key point is that if there is no persistent increase in head, then there has been no increase in the volume of stored water. Recharged water may leak out of the system, be lost to discharge, or abstracted by other users who may not be participants in the groundwater banking system. Additionally, where the area of an aquifer is very large (e.g., a regional aquifer), the increase in head for a given volume of recharge becomes imperceptible. Another fundamental aspect of the hydrology of groundwater banking systems is the difference between dynamic and static responses to recharge and recovery (abstractions). During recharge and recovery, water levels will correspondingly rise or fall in the vicinity of the wells or infiltration

4 R. G. Maliva / Water Policy 16 (2014) basins. This dynamic change in heads quickly dissipates once recharge or recovery stops. Water levels return towards background levels. For groundwater banking systems, the key issue is the change in static water levels that persists after the dynamic response from recharge or recovery dissipates (Figure 1). If water levels recover to pre-recharge levels (Δh ¼ 0), then no local change in storage has occurred. A consequence of the difference between the dynamic and static response is that the small or moderate rise in static water levels resulting from recharge will only partially offset the larger local dynamic drawdowns from later pumping. Where local hydrologic impacts (e.g., maintenance of spring or river flows) are a constraint on groundwater use, then the location and timing of both recharge and recovery are critical issues for the operation of groundwater banking systems. Aquifer recharge can be performed by either injection or land application (surface spreading). For land application systems, such as infiltration basins, some applied water will not reach the water table, particularly where the water table is deep, significant lateral flow occurs in the vadose zone, and/or there is a high rate of evaporation relative to the infiltration rate. Actual aquifer recharge in land application systems, especially in arid climates, may be significantly less than the volume of water applied. Water can also be banked by in-lieu recharge, whereby credit is given for allocated groundwater that is not withdrawn. In terms of an aquifer water budget, the net effects of recharging a given volume of water or not withdrawing the same volume of water that would otherwise be used are similar. In-lieu recharge may be more effective than direct groundwater recharge because it does not have associated water losses, such as percolating water that does not reach the water table (Franson, 1989). In-lieu recharge may involve the substitution of surface water for groundwater or a reduction in actual water use (e.g., reduction in irrigated area or adoption of more water-efficient irrigation techniques). Although the in-lieu recharge concept is technically sound, it can be subject to abuse if credits are issued for reductions in groundwater use that would normally have occurred anyway, and there are thus no real savings to an aquifer. Groundwater users often do not pump their full allocation each year because the water is not needed. For example, during a wetter than normal growing season, farmers will usually irrigate less. Allowing credits to be earned for reduced groundwater pumping in wet years, Fig. 1. Conceptual diagram illustrating dynamic and static responses to injection and recovery. Large dynamic changes in water levels may occur near the sites of injection and recovery, which dissipate once injection or recovery is terminated. Injection results in net storage of water only if there is a persistent increase in heads (Δh) after the injection is terminated (track A). Injection results in no net storage if water levels recover to static levels (track B).

5 148 R. G. Maliva / Water Policy 16 (2014) which will later be used for actual groundwater pumping during drier years, will have the effect of increasing actual groundwater use. This will cause a negative impact on the aquifer water budget. An important technical issue is the aquifer response to both managed aquifer recharge and recovery. Where an aquifer is hydraulically connected to a surface-water body, increases in the water table elevation as the result of managed recharge may either increase the rate of discharge in gaining stream reaches or decrease the amount of induced recharge in losing stream reaches, both of which cause water to leave the basin and reduce the potential increase in aquifer storage (Purkey et al., 1998; Thomas, 2001; Contor, 2009). Managed aquifer recharge can cause stream flow to be larger than would otherwise occur. The opposite would occur during recovery, where the local lowering of the water table could increase the rate of infiltration, thereby reducing stream flow. The recharge and recovery rate of water in groundwater banking systems could also affect the rates of lateral and vertical groundwater flow into an aquifer. Therefore, a one-to-one correspondence may not occur between the volume of recharged or recovered water and the change in aquifer storage, which needs to be considered and incorporated into the groundwater bank accounting system. 3. Water accounting system The main objectives of the water bank accounting and regulatory system are to track water deposits and withdrawals and to control the amount, timing, and location of withdrawals by the participants in the system and other aquifer users. Where groundwater use is not closely regulated, the opportunity exists for unauthorized water users to free-ride and take advantage of the managed recharged water. The water accounting system for groundwater banks serves to: protect the rights of bank members, and ensure that they receive benefits commensurate with their deposits; optimize the economic productivity of water by facilitating transfer of water credits; ensure the sustainability of the system by preventing (or mitigating) adverse impacts associated with operation of the system. Contor (2009) proposed that the double-entry accounting method be used, in which every transaction is recorded as both a debit entry and a credit entry in separate ledger accounts, and at all times the accounting system tracks both inventory in asset accounts and claims to inventory in ownership accounts. Deposit activities are actions that cause more water to be stored in an aquifer than would otherwise have been the case (Contor, 2009). With respect to the operation of a groundwater banking system, the recharge of a given volume of water by a system participant would be entered as a credit in the participant s account and as a liability (i.e., water owed) in the groundwater bank s account. The basic requirement for a double-entry accounting, and the successful operation of a groundwater bank, is that both ledgers be balanced. The assets (i.e., rights to water) held in the participants accounts should be equal to (or less than) the volume of water that can be physically recovered from the system. If assets are greater than liabilities, then the groundwater bank is insolvent, although the insolvency may not be manifest until a drought occurs when multiple users attempt to withdraw water. As is the case for a financial bank, an unsound groundwater bank may continue to operate for a long time as long as

6 R. G. Maliva / Water Policy 16 (2014) annual deposits are greater than withdrawals. The insolvency might be revealed when users try to withdraw a greater number of credits during a drought than the aquifer can safely provide. Despite its importance, the capability of groundwater banks to provide accumulated credits has seldom been publicly addressed. Data are available for the Las Posas Basin (Southern California) groundwater bank, which provide an example of the types of issues that may arise (Maliva & Missimer, 2010). Fox Canyon Groundwater Management Agency (FCGMA), the regulatory agency that has jurisdiction over the project, noted that the historic accumulation of credits within the FCGMA has been steadily increasing, approaching 550,000 acre-feet (AF) ( m 3 ) in 2006 (Fox Canyon Groundwater Management Agency, 2007). The estimated total net credit balance in the East, West, and South Las Posas Basin at the end of calendar year 2006 was 116,002 AF ( m 3 ) compared to an annual abstraction of 27,234 AF ( m 3 ). The accumulated credits are over four times the annual abstraction rate. The volume of credits that are accumulating through the operation of the recharge system and in-lieu recharge greatly exceeds the amount of water that could be extracted during a short-time period (e.g., major drought). The FCGMA (2007) noted that the use of a significant number of credits in a short period of time during a period with limited groundwater recharge represents a threat to the regional groundwater resource. It was noted that even a 5% use of the total amount of credits currently available would result in a net 24% increase in annual extraction, which could result in persistent depressions in groundwater elevations, land subsidence, and seawater intrusion (Fox Canyon Groundwater Management Agency, 2007). The US$150-million-dollar project is now recognized to be a failure that was marred by insufficient research, poor judgment and hollow assurances (Blood & Spagat, 2013). The accounting system can provide opportunities for increasing the efficiency of water use where the credits for stored water can be sold, purchased, or traded. For example, a party with excess surface water might use it to recharge an aquifer and later sell the accumulated credits to another party with a greater need for the water (Frederick, 1995; Contor, 2009). Groundwater systems can thus facilitate the reallocation of water to higher-value uses, provided that any regulatory obstacles to the practice can be overcome. 4. Regulatory framework Groundwater banking systems also require regulatory or other control over the groundwater basin, to enforce the design withdrawal rates of system participants and to exclude other aquifer users from initiating or increasing abstractions and thus taking the stored water. Under ideal circumstances, the groundwater bank owner or its participants are the sole groundwater users in the basin, and abstractions can be readily controlled. Otherwise, some mechanism must be in place to control the actions of the other users such as a regulatory restriction against additional abstractions. Total abstractions may be constrained during some periods by external considerations, for example, a regulatory requirement to maintain in-stream flows for ecosystem or species protection. In these situations, the decision needs to be made as to who has priority to the stored water. For example, do the rights of groundwater bank participants to stored water supersede the rights of more-senior existing groundwater users? Purkey et al. (1998) and Thomas (2001) discuss the technical and regulatory issues associated with groundwater banking in the state of California, USA. Major regulatory and organizational challenges

7 150 R. G. Maliva / Water Policy 16 (2014) may occur when multiple entities have jurisdiction over all or parts of a project. For example, in California, different entities may be involved including (Thomas, 2001): reservoir owner who is in charge of the storage and release system used to generate source water for groundwater banking; local groundwater management authority that would allow for the rent of aquifer space for temporary storage; operator of infrastructure needed to move water from the reservoir to the groundwater bank and then to the point of use; end users who would pay for the new yield and generate revenue streams to compensate the reservoir owner and groundwater banker; regulatory agencies with jurisdiction over groundwater and surface-water use; regulatory agencies with jurisdiction over groundwater and surface-water quality; and regulatory agencies with jurisdiction over environmental-protection issues. While many of the regulatory issues are state-specific, the general concepts are broadly applicable. A basic requirement for any water banking scheme is that some mechanism must be in place to prevent the stored water from being abstracted by other aquifer users, particularly those who are not participating in the system. The accounting and regulatory framework for groundwater banking systems can be as important as, and must rely on, a detailed knowledge of local hydrogeology in determining the longterm success of the systems. The systems need to be operated so that they provide commensurate benefits to justify construction and operation of the system. 5. Groundwater modeling The impacts of the operation of a groundwater banking system on aquifer water levels were simulated using a numerical model of a hypothetical groundwater basin using the MODFLOW code (McDonald & Harbaugh, 1988). The simulated basin has an area of 10 km by 20 km, saturated thickness of 100 m, horizontal and vertical hydraulic conductivity of 50 m/d, and specific yield of 0.1. The model cells are 100 m by 100 m, and the unconfined aquifer is simulated as being surrounded by no flow cells. The groundwater banking system is simulated as centralized vertical alignment of ten wells with a spacing of 100 m and injection and abstractions rates of 50,000 m 3 /day. A groundwater banking system operated for only seasonal recovery of water was simulated by injecting for four months, followed by two months of storage, four months of abstraction, and then two months of no activity to complete the annual cycle (Figure 2(a)). Water levels in the immediate vicinity of the groundwater banking system wells are dominated by transient dynamic responses. More distant (3 km) from the wellfield, the dynamic response becomes muted relative to the static response. Owing to the relatively low hydraulic diffusivity (transmissivity divided by storativity) of the simulated unconfined aquifer, local water levels do not completely recover back to static level after injection and abstractions during the 60-day storage or rest periods. The low hydraulic diffusivity also results in peaks and troughs in water levels occurring later and at increasing distances from the wellfield. The offset is approximately 30 days at a 3 km distance in the simulated scenario.

8 R. G. Maliva / Water Policy 16 (2014) Fig. 2. Hydrographs from model of a hypothetical groundwater bank. (a) Simulation of equal seasonal injection and recovery. Hydrograph from center of wellfield (solid line) reflects large local dynamic responses. Hydrograph for a well 3 km away from wellfield (dashed line) shows a much muted response to operation of the system. Water level highs and lows are offset by about 30 days. (b) Simulation of 8 years of seasonal recharge followed by recovery of all stored water over two seasons of drought. The hydrograph from wellfield (solid line) shows a dynamic response from injection and very large local drawdowns during recovery. Distant (3 km) well hydrograph (dashed line) shows a rise in static water level during injection period and recovery back to near static levels after recovery. (c) Simulation of a groundwater banking system in an aquifer experiencing overdraft. During the simulated 8 years of recharge, the recharged water balanced the overdraft as indicated by stable water levels in the distant (3 km) well hydrograph (dashed line). During the 2 years of simulated seasonal recovery of the recharged water, large dynamic drawdowns occur in the wellfield (solid line) and static water levels drop as a result of the previous overdraft, which was hitherto not evident.

9 152 R. G. Maliva / Water Policy 16 (2014) Groundwater banking for long-term storage was simulated as 8 years of seasonal recharge, followed by seasonal abstraction at four times the injected rate for 2 years (i.e., complete recovery of the injected water). This scenario would occur where excess water is banked during wet or normal years and then recovered at a high rate during drought periods. Seasonal recharge is simulated to continue thereafter. The hydrographs for the wellfield at the 3 km location reflect both the dynamic responses to pumping and an increase in static water levels of about 2 m (Figure 2(b)). The abstraction of the recovered water resulted in large simulated drawdowns near the wellfield and a lesser, abrupt drop in static water levels. Aquifer overdraft was simulated by applying a negative recharge of m/d to the long-term storage scenario (Figure 2(c)). A key feature of the simulation results is that the net recharge from a groundwater banking system can mask the aquifer overdraft from other aquifer users. The managed recharge compensated for the overdraft and temporarily resulted in stable static water levels. However, once the abstractions start, large simulated dynamic drawdowns occur near the wellfield, and static water levels abruptly drop, revealing the effects of the historic overdraft on aquifer static water levels. The magnitude of the dynamic and static responses depends upon the volume of recharged water, recharge and abstraction rates, aquifer hydraulic properties, and aquifer area. 6. Discussion Successful operation of groundwater banking systems requires that most of the recharged water is recoverable by system participants when needed. A major threat to sustainable operation of groundwater banking systems is that accumulated credits may eventually exceed the volume of water that can be safely produced during a high-demand period. This challenge is exacerbated by the likelihood that withdrawals from the groundwater bank may be concentrated during drought periods when aquifer water levels may be relatively low and demand is high. Local dynamic drawdowns during recovery can impact local sensitive environments and other aquifer users. Aquifer-wide increases in static water level will not offset the much larger dynamic drawdowns that may occur during local abstractions. In regional confined aquifers, recharge may not result in any material increases in local aquifer pressures that persist until the time of abstractions. For example, the San Antonio Water System (SAWS) Twin Oaks ASR creates a cone of depression in the Carrizo Aquifer during recovery that is similar to that of a purely extractive wellfield. Impacts during recovery were recognized and the Carrizo Aquifer Well Mitigation Program was implemented under an inter-local agreement even though mitigation was not required under Texas water law (Evergreen Underground Water Conservation District, 2006). The well mitigation process may include lowering of pumps, drilling of replacement wells, or connection to an existing water purveyor. Similarly, recovery of water from water banks in Kern County, Southern California, during a drought period, was reported to have adversely impacted local water users (Barringer, 2011). Sustainable groundwater banking requires that accumulated credits be kept in line with the safe aquifer yield. The term safe yield is referred to herein as the volume of water that can be abstracted from a groundwater banking system at a given time without causing unacceptable adverse impacts. Sustainable operation of the systems thus requires an accurate understanding of the aquifer or basin water budget through both monitoring and numerical modeling. It is critical to quantify the main water budget elements and determine how much water can safely be withdrawn during various time periods. Some parameters (R N, D, L) are difficult to measure directly and may have to be estimated through the

10 R. G. Maliva / Water Policy 16 (2014) model calibration process. Groundwater abstractions by all users, managed recharge, and aquifer water levels need to be measured and stored (ideally in real time) in a central database. Several solutions are available to prevent accumulated credits from exceeding the aquifer safe yield. Where some of the recharged water is lost by leakage or other means, a discount could be applied to withdrawals (Contor, 2009, 2010). Less than 100% of the recharged water may be allowed to be recovered. Commonly, a 10% loss is assumed in California groundwater banking systems, which is subject to adjustment based on monitoring results (Thomas, 2001). Groundwater losses could be subtracted from the participants bank accounts proportional to the amount of water stored in each account (Sandoval- Solis et al., 2011). The discount could be applied once, at the time of recharge, or there could be periodic (e.g., yearly) discounting of credits. When it is recognized that the accumulated credits exceed the safe yield of a system, all credits can be discounted (devalued) to bring the system back into balance. For example, a factor of 0.9 could be applied to all the credits to address a 10% imbalance in the system. Credits for earlier recharged water would undergo numerous devaluations over time, which creates a disincentive for long-term hoarding of water. Periodic discounting maintains incentives for new recharge intended for short- and intermediate-term use. However, when the discount applied to recharge credits is too high, then there can be little economic incentive to store water. Credits can be given a finite life and expire after a specified time if not used. Short credit lives (e.g., 1 year) minimize the risk of an imbalance between accumulated credits and system safe yield, but reduce the value of the system to participants. An intermediate-length life (e.g., 5 to 20 years) would increase the drought-protection benefits while still reducing potential imbalances. The amount of credits that can be recovered in a given year can be restricted to an annual safe yield volume. Inasmuch as the recoverable credits will be less than the accumulated credits, a system would be needed to allocate the recoverable credits among system participants. One option is to prorate the recoverable credits amongst all the credit holders based on the number of credits held. The drawback is that accumulated credits would have lesser value and more uncertainty as a drought-proofing tool than participants may have expected. Limiting annual abstractions would also not address the problem of a progressively increasing number of accumulated credits in the bank. From a physical perspective, if recharged water is never abstracted, it will be discharged from the banking system by natural groundwater flow. The physical system must be accurately understood and represented in a groundwater-flow model, so that administrative credit devaluation accurately represents natural credit devaluation. Additional options are either to accept the adverse impacts as a cost of the system or to mitigate the impacts. The owner or operator of the system could be required (or volunteer) either to compensate aquifer users impacted by the banking system rules or to implement mitigation measures for the natural system. These decisions also require an accurate representation of the physical system in a groundwater-flow model. The timing and location of withdrawals also need to be considered for sustainable groundwater banking systems. Groundwater abstractions result in local aquifer drawdowns, which can have adverse impacts, such as reductions in stream and spring flows and wetland water levels, and the lowering of water levels in wells. Local adverse impacts can occur even though the system is neutral in terms of the overall aquifer water budget. For example, if maintenance of dryseason spring flows is a limiting factor, then wet-season recharge may not offset additional dryseason abstractions in the vicinity of the springs. Contor (2009, 2010) proposed the use of surface-water aquifer-response functions that would equalize the hydrologic values of recharge and abstraction with respect to time and location. Groundwater

11 154 R. G. Maliva / Water Policy 16 (2014) modeling can be employed to assess the impacts of proposed recharge and subsequent withdrawals on surface-water bodies and the aquifer. Returning to the spring example, the credits granted for recharge would depend upon the degree to which the recharge is demonstrated through modeling to result in an increase in spring flow during the time periods of concern. Similarly, the number of credits required to abstract a given volume of water would depend upon the modeled impacts of the abstractions on spring flows during periods of concern. Depending upon local circumstances, a substantial discount may have to be applied to avoid adverse impacts if they are to occur during dry periods when surface-water bodies are most vulnerable. From an economic perspective, the use of aquifer-response functions turns water into a homogenous quantity because the withdrawal point and time no longer make a difference with respect to surface-water impacts (Contor, 2010). The lack of a link between the banking system and the limitations of the physical system can be a cause of system breakdown. The management challenges of groundwater banking systems become even more complicated when there are numerous aquifer users and many of the users do not participate in the groundwater banking system. All aquifer users may benefit from the net recharge in terms of higher aquifer water levels, which may encourage additional groundwater use. Existing groundwater users may benefit from higher groundwater levels and thus lower pumping costs while water is stored (Davis et al., 2001). For example, a cost-benefit analysis of the Las Vegas Valley Water District (LVVWD) artificial recharge system in southern Nevada demonstrated that the overall benefits of the artificial recharge system are greater than the costs. Operation of the system benefits all aquifer users by lowering energy costs, decreasing the need to deepen wells, lessening impacts from land subsidence, and providing additional water for the aquifer system (Katzer et al., 1998; Donovan et al., 2002). The nonmunicipal aquifer users were receiving free benefits from the system, as they were not paying towards the system operation. The solution adopted to this free-rider situation was to bill an annual groundwater management fee to well owners and groundwater permit holders in the Las Vegas basin to support the artificial recharge program and other groundwater management and protection programs. 7. Conclusions Where local hydrogeological conditions are favorable for the practice, groundwater banking can be a valuable tool for IWRM by providing a means for storing excess surface-water flows that might otherwise not be put to beneficial use. However, operation of groundwater banking systems requires careful planning in order for the systems to be sustainable in terms of being able to recover stored water when needed, while avoiding unacceptable adverse impacts to the environment, surface waters, and other aquifer users. The key challenge is developing an operational/administrative scheme that is based on detailed physical knowledge, and accurate numerical representation, of the aquifer or groundwater basin water budget. The scheme must adequately address the temporal and spatial differences in (and dynamic responses to) recharge and abstractions. Even though the operation of a groundwater bank may be neutral or beneficial in terms of the aquifer water budget (i.e., result in a net aquifer recharge), the operation of a system may still cause significant local adverse impacts. An important component of sustainable groundwater banking is an accurate groundwater model of the aquifer or basin that can be used to quantitatively evaluate the impacts of both recharge and abstractions, which are then used to refine the water accounting system of the bank. The main challenge to sustainable long-term groundwater banking systems is ensuring that credits accumulated to withdraw from the bank are accurately

12 R. G. Maliva / Water Policy 16 (2014) discounted as necessary over time, to be consistent with the ability of the aquifer to provide the water without adverse impacts. Acknowledgments I would like to thank Mike Geddis, Kathy Champagne-Baker, and three anonymous Water Policy reviewers for their thoughtful reviews. References Barringer, F. (2011). Storing Water for a Dry Day Leads to Suits. New York Times, July 27, Blood, M. R. & Spagat, E. (2013). Las Posas Basin Aquifer Failure Illustrates Risks of Underground Reservoirs. Associated Press, August 24, Bredehoeft, J. D. & Young, R. A. (1983). Conjunctive use of groundwater and surface water for irrigated agriculture: risk aversion. Water Resources Research 19(5), Clifford, P., Landry, C. & Larsen-Hayden, A. (2004). Analysis of Water Banks in the Western States. Washington Department of Ecology. Publication No , Olympia, WA. Contor, B. A. (2009). Groundwater banking in aquifers that interact with surface water: aquifer response functions and doubleentry accounting. Journal American Water Resources Association 45(6), Contor, B. A. (2010). Status of ground water banking in Idaho. Journal of Contemporary Water Research & Education 144(1), Davis, M. D., Lund, J. R. & Howitt, R. E. (2001). Banking groundwater to meet growing water demand in California. In: Bridging the Gap: Meeting the World s Water and Environmental Resources Challenges, Proceedings of World Water and Environmental Resources Congress 2001, p. 13. Dillon, P. (2005). Future management of aquifer recharge. Hydrogeology Journal 13(1), Donovan, D. J., Katzer, T., Brothers, K., Cole, E. & Johnson, M. (2002). Cost benefit analysis of artificial recharge in Las Vegas Valley, Nevada. Journal Water Resources Planning and Management 128(5), Evergreen Underground Water Conservation District (2006). San Antonio Water System s Twin Oaks Aquifer Storage & Recovery Program Status. PowerPoint presentation (January 16, 2006). Fox Canyon Groundwater Management Agency (2007). Annual Report for Calendar Year Fox Canyon Groundwater Management Agency, Ventura, CA. Franson, J. W. (1989). Evaluating potential artificial recharge projects. In: Artificial Recharge: Proceedings of the International Symposium: Anaheim, California, August 22 27, Johnson, A. I. & Finlayson, D. J. (eds). American Society of Civil Engineers, Reston, VA, pp Frederick, K. D. (1995). Adapting to climate impacts on the supply and demand of water. Climatic Change 37(1), Greydanus, H. W. (1978). Management aspects of cyclic storage of water in aquifer systems. Journal American Water Resources Association 14(2), Harou, J. J. & Lund, J. R. (2008). Ending groundwater overdraft in hydrologic-economic systems. Hydrogeology Journal 16(6), Katzer, T., Brothers, K., Cole, E., Donovan, D. & Johnson, M. (1998). A Cost-Benefit Analysis of Artificial Recharge in the Las Vegas Valley Ground-Water System, Clark County, Nevada. Report prepared for the Southern Nevada Water Authority Ground Water Management Program, p. 38. Maliva, R. G. & Missimer, T. M. (2008). ASR, useful storage, and the myth of residual pressure. Ground Water 46(2), 171. Maliva, R. G. & Missimer, T. M. (2010). Aquifer Storage and Recovery and Managed Aquifer Recharge using Wells: Planning, Hydrogeology, Design, and Operation. Schlumberger Water Services, Methods in Water Resources Evaluation Series No. 2, p McDonald, M. G. & Harbaugh, A. W. (1988). MODFLOW. A Modular Three-Dimensional Finite-Difference Ground Water Flow Model. U.S. Geological Survey Techniques of Water-Resources Investigation Report 06-A1, p. 586.

13 156 R. G. Maliva / Water Policy 16 (2014) Miller, K. A. (2000). Managing water supply variability: the use of water banks in the western United States. Drought: A Global Assessment, Volume II. Wilhite, D. A. (ed.). Routledge, London, pp Pulido-Velazquez, M., Jenkins, M. W. & Lund, J. R. (2004). Economic values for conjunctive use and water banking in southern California. Water Resources Research 40, W Purkey, D. R., Thomas, G. A., Fullerton, D. K., Moench, M. & Axelrad, L. (1998). Feasibility Study of Maximal Program of Groundwater Banking. Natural Heritage Institute, San Francisco, CA, p. 77. Sandoval-Solis, S., McKinney, D. C., Teasley, R. L. & Patino-Gomez, C. (2011). Groundwater banking in the Rio Grande Basin. Journal of Water Resource Planning and Management 137(1), Thomas, G. A. (2001). Designing Successful Groundwater Banking Programs in the Central Valley: Lessons from Experience. The Natural Heritage Institute, Berkeley, CA, p Thomas, H. E. (1978). Cyclic storage, where are you now? Ground Water 16(1), Received 31 January 2013; accepted in revised form 18 July Available online 30 October 2013