New Economic Paradigm for Sustainable Reservoir Sediment Management

Reservoir planning and economic studies are evaluated over a specified future time period that is called POA. In the planning phase of a new water storage project a BCA is conducted that evaluates a selection of benefits and costs over this POA. It typically covers either 50 or 100 years (Morris and Fan 2010). The concept that infrastructure will serve its purpose for a finite period is called design life (Annandale et al. 2016). If well maintained, dam structures may last centuries. However, sedimentation will impact the operations of dam and reservoir facilities long before the reservoir completely fills with sediment. Decommissioning of high hazard dams will often be necessary after reservoir operations become significantly impaired (Randle et al. 2021).

The reservoir storage reduction due to sedimentation is a relatively slow process (Huffaker and Hotchkiss 2006) and produces a low rate of benefit loss (Coker et al. 2009). Sedimentation impacts along the upstream channel, and degradation impacts along the downstream channel, tend to be experienced more rapidly, but the economics of those impacts have not been considered. Therefore, traditional applications of BCA in water project planning do not comprehensively account for the costs of sedimentation, and consequently conclude that additional capital costs to manage sedimentation are not economically justified (Kondolf et al. 2014). The design life approach analyzes reservoir benefits and costs over a certain time period and does not treat water storage as a resource in perpetuity (Johndrow et al. 2006). The historical approach neglected the loss of water storage over time, the eventual cost of dam decommissioning, and damages from upstream sedimentation and downstream erosion. A reason for this historical approach might be that the present value of annual costs is seldom significant beyond 50 years because they are heavily discounted. However, reservoir storage sites are an exhaustible resource, so the design life approach will likely leave future generations with few and expensive options to consider. At the end of reservoir life, and eventual dam decommissioning, a replacement project would be expected to at least maintain the historical benefits that were provided by the previous reservoir. The replacement dam and reservoir project will require a new economic analysis and justification. However, new reservoir sites will not be as plentiful and will impose greater engineering and permitting challenges. For example, in the 1970s the Denver Board of Water Commissioners proposed building Two Forks Dam to help meet the water supply needs of the Denver metropolitan area. After 20 years of planning, the permit to construct the dam was denied by the US Environmental Protection Agency (USEPA 1990).

Dams are built mainly to provide an irrigation and municipal water supply; flood risk reduction; recreation, fish and wildlife benefits; hydropower; and river navigation. However, they entail huge investment costs, such as planning and design, construction, land purchase, and resettlement, and involve other important factors, such as social and environmental impacts. The latter can be estimated using different methods, such as found in Chutubtim (2001), Dunning and Durden (2009), Durden and Johnson (2013), and Zerbe and Scott (2015). Knowing as much as possible about the costs and the benefits leads to better decisions (Ackerman 2008). Historically, economic analyses did not fully capture all the temporal, spatial, environmental, and social dimensions of dam construction projects. The positive and negative externalities, such as upstream and downstream environmental damages, infrastructure and land use changes, water quality, and flood stage, received very little attention. Given what is known today, economic assessments of water projects should be modified to consider all anticipated costs and benefits. Without such considerations, any analysis is incomplete and therefore faulty (Hotchkiss and Bollman 1996).

Discounting is a mathematical procedure employed to make costs and benefits, which occur at different points in time, temporally equivalent. Discounting for temporal equivalence can be achieved using a variety of different approaches, referred to collectively throughout this paper as a discounting approach. The choice of discounting approach and discount rate have a high impact on the net present value (NPV) and the BCR of projects with a significant difference in the timing of costs and benefits. Projects like dams, which require large initial capital outlays and benefits distributed across the lifetime of the project, are greatly affected by the discounting approach and rate (Congressional Research Service 2016; National Center for Environmental Economics 2014).

To reflect any serious responsibility to future generations, the discount rate must be quite low (Ackerman 2008; Weitzman 1998), but not zero. Zero discounting means current generations should reduce their incomes to benefit future generations (Harpman and Piper 2014), or sacrifice the current generations’ (the poorest generation) well-being (Pearce et al. 2003) relative to future generations. A recently proposed solution to this problem is to use a discount rate that declines with time to raise the weight attached to the welfare and well-being of future generations (Groom et al. 2005). In recent years new discounting approaches have attracted attention, but there is no agreement among economists regarding their use in water resource economic assessments (Harpman and Piper 2014).

Developing and retaining enough reservoir storage space to satisfy water demand over the very long term requires abandoning the conventional design-life approach to dam design and adopting a life-cycle management approach. A major difference between the life-cycle management approach and the design-life approach is the use of sediment management to preserve reservoir storage capacity over time (Annandale et al. 2016). The life cycle approach offers a framework for sustainably maintaining project benefits across generations (Palmieri et al. 2003). When sedimentation is controlled, dams can have useful lives exceeding any other type of engineered infrastructure (Morris et al. 2008), to meet current and future generations’ water demands. The authors are not aware of any previous study that comprehensively compares these two approaches for new and existing reservoirs.

Johndrow et al. (2006) estimated that an annual investment of between $10–$20 billion would be required for the construction of replacement dams and reservoirs to recover current worldwide reservoir storage loss due to sedimentation, without additional storage creation. Replacing lost reservoir storage capacity by constructing new dams would be challenging due to high land prices and a lack of favorable sites. Even when technically feasible sites exist for new dams, they may not be feasible from an economic, social, political, or environmental standpoint. Even before the major dam-building decades in the US, Brown (1946) recognized that major reservoirs are irreplaceable when he said:

If the contemplated public and private reservoir construction programs are carried out, we shall have utilized by the end of this generation a very substantial portion of all the major reservoir sites … We cannot discover new reserves, as we will of oil. Nor we can grow new resources, as we can of forests. To whatever degree we conserve the capacity of the reservoirs built on these sites, to just that degree shall we conserve this indispensable base of our national strength and prosperity.

Reservoir sediment management can be considered for new and some existing dams, although (for example) retrofitting low-level outlets for sediment management is more expensive than incorporating such outlets in the initial design and construction. The preserved benefits from sustainable sediment management can offset the additional costs, as will be demonstrated in the case study portion of this paper. In the absence of sediment management, those economic benefits that are dependent on reservoir capacity are progressively reduced as sedimentation increases. These lost benefits should be accounted for in an objective economic analysis (Song et al. 2021).

Unsustainable reservoir sedimentation not only causes storage capacity loss, but also leads to upstream and downstream environmental, social, cultural, and economic impacts. Upstream and within the reservoir, sediment accumulation will eventually bury dam and reservoir facilities, reduce recreation use, and impact upstream property through increased flood stage and groundwater levels (George et al. 2016). The downstream sediment deficit results in the erosion of the stream channel beds and banks, disrupts natural ecosystems and channel substrate compositions, threatens riverine infrastructure, and leads to coastal delta and beach erosion (Kondolf et al. 2014). Thus, the footprint defined for economic evaluation should reach far enough up- and downstream to capture all sedimentation or sediment-deficit impacts comprehensively.

For new reservoirs, implementing adequate spatial and temporal scales is essential. Objective economic analysis requires a long view of future effects as well as a broad assessment of impacts. Comprehensive accounting of benefits and costs over a sufficiently long POA will yield a BCR that fairly evaluates sustainable development. Such an analysis could indicate whether sustainable sediment management is economically justified, given the cost of sedimentation impacts and eventual dam decommissioning.

For an existing dam and reservoir, there are two categories of alternatives: sustainable sediment management that maintains the remaining storage capacity indefinitely, and non-sustainable sedimentation that requires eventual dam decommissioning. An objective economic comparison of these alternatives requires a comprehensive accounting of all cost and benefits that can be compared by NPV. NPV requires the same inputs as BCA, but outputs a metric useful for identifying maximum net benefits. In the absence of sediment management, economic benefits dependent on reservoir capacity are progressively reduced as sedimentation increases. These lost benefits should be accounted for in an objective economic analysis (Song et al. 2021).

BCA is especially useful in determining the economic feasibility of an action alternative in comparison to a no action alternative—For example, determining the economic feasibility of a new dam construction project, where no action means not constructing the project. However, when considering sedimentation of existing reservoirs, no action eventually results in the decommissioning of the dam and lost project benefits. NPV provides a more intuitive and meaningful metric for comparing alternatives for existing projects, allowing decision makers to identify the alternative that maximizes net economic benefits.

Exponential (classic) discounting is the approach traditionally used by economists and engineers. When exponential discounting is employed, costs and benefits occurring several decades into the future, even dam decommissioning cost, have reduced influence on the initial investment decision. Projects can be economically justified without sediment management, but this can lead to intergenerational inequity. Several new discounting approaches have been described in recent years. Fig. 1 illustrates the temporal differences across a selection of discounting approaches over a 150-year period. Arguably, the new discounting approaches may better represent future economic uncertainty, regional and intergenerational equity, and sustainability considerations (Harpman and Piper 2014). As a group, these new discounting approaches may be better suited than the exponential discounting approach for the analysis of long-lived infrastructure and environmental investments. Many of these new discounting approaches result in declining discount rates (DDRs) over time. DDRs may be more appropriate than constant discounting for reservoir economic assessment, climate change issues, or other projects with intergenerational dimensions (National Center for Environmental Economics 2014). DDRs have also been used by World Bank Group projects (Annandale et al. 2016).

Among the nine discounting approaches depicted in Fig. 1, three are investigated in the case study section: exponential, hyperbolic, and inter-generational. The equations expressing these discounting approaches, as well as a brief description of each, are provided subsequently.

Exponential discounting:where ${\mathrm{w}}_{\mathrm{t}}$ is the discount factor or weight at time (t); r is the (constant) discount rate; and t is a time period index.

Discounting future benefits or costs by a fixed rate, for each unit of time, is the basis of exponential discounting. However, as mentioned earlier it can be problematic and inappropriate for investments that are to be judged over longer periods of time, because future generations will bear costs (or benefits) from actions of previous generations (Guerriero and Pacelli 2020).

More generally, the rate at which people discount future benefits and costs declines as the length of the delay increases (Redden 2007). Hyperbolic discounting is an alternative discounting approach that decreases the rate of discounting as the delay occurs further in the future. Hyperbolic discounting will generally discount future benefits and costs more than exponential discounting for short delays, and less than exponential discounting for long delays (Redden 2007).

Hyperbolic discounting:where ${\mathrm{w}}_{\mathrm{t}}$ is the discount factor or weight at time (t).

The parameter $h>0$ controls the effect of time perception, and $k>0$ influences the degree to which the hyperbolic discount factor differs from exponential discounting (Harpman and Piper 2014).

Preferences can change over time, and this characteristic makes it difficult for analysts to assess whether current generations’ preferences reflect those of communities that are not born yet (Guerriero and Pacelli 2020). An alternate method of incorporating intergenerational impact is to consider the timespan of future generations. Intergenerational discounting accomplishes this by requiring two different discount rates and an assumed generation timespan (Harpman and Piper 2014; Sumaila and Walters 2005).

Intergenerational discounting:where ${\mathrm{w}}_{\mathrm{t}}$ is the discount factor or weight at time (t); ${\mathrm{r}}_{\mathrm{a}}$ is the present generation annual discount rate; ${\mathrm{r}}_{\mathrm{fg}}$ is the future generation annual discount rate; G is the assumed generation timespan; and t is a time period index, and

San Clemente Dam near Carmel, CA filled with sediment and was decommissioned in 2015 due to dam safety and environmental concerns and lack of project benefits. The dam was completed in 1921 and by 2008 was providing less than 5% of its original storage capacity, at which time California American Water (dam owner) had to ask the California Public Utilities Commission for a rate increase to pay its share of the dam removal cost ($49 million). The current generation (notably rate payers) paid for the entire dam decommissioning cost but received little or none of the water storage benefits (California Public Utilities Commission 2012).

For existing reservoirs, some actions will have to be taken. In 2017, the Federal Advisory Committee on Water Information and its Subcommittee on Sedimentation approved a resolution on Reservoir Sustainability (ACWI 2021). This resolution asked all Federal agencies to “develop long-term reservoir sediment-management plans for the reservoirs that they own or manage by 2030. These management plans should include either the implementation of sustainable sediment-management practices or eventual retirement of the reservoir.” In 2018, the US Society on Dams adopted a similar resolution for owners of all dams and reservoirs (USSD 2018).

One mechanism for financing eventual dam decommissioning is through a decommissioning fund. Such a fund would have a maturation date equal to the expected dam decommissioning year, based on the sedimentation rate and other assumptions. Annual contributions to the fund would be paid by project beneficiaries and would be calculated as the cost of decommissioning (in present dollars) amortized over the remaining years of dam life. The decommissioning fund will approximate the cost of decommissioning in the year of dam removal. This fund could also serve to offset the cost of any emergency actions required as the dam and reservoir age.

For example, federal water projects are authorized for specific project beneficiaries who are responsible for a portion of project repayment through a cost allocation framework based on the “beneficiary pays” principle (Congressional Research Service 2020). Some portions of the project benefits (e.g., flood risk reduction, recreation, and fish and wildlife habitat) are often assigned to the American public. This framework could be extended to dam decommissioning fund annual contributions. Establishment of a dam decommissioning fund could help achieve intergenerational equity to prevent future generations from having to pay for dam decommissioning when they receive little or no project benefits.

The concept of a dam decommissioning fund is similar to that used in natural resources extraction practices. The Surface Mining Control and Reclamation Act of 1977 (SMCRA) provides that, as a prerequisite for obtaining a coal mining permit, a permittee must post a reclamation bond to ensure that the regulatory authority has sufficient funds to reclaim the site in case the permittee fails to complete the approved reclamation plan (US Department of the Interior 2021).

An additional consideration is the comparison of the annualized cost of sustainable sediment management to the annual contribution required to a decommissioning fund. If the comparison indicates that the sustainable sediment management cost less than decommissioning fund contributions, this bolsters the economic case for sediment management. This comparison also presents an additional way to conceptually present objective economic analysis to key decision-makers.