Review of the U.S. Army Corps of Engineers Restructured Upper Mississippi River-Illinois Waterway Feasibility Study: Second Report (2004)

This is the second of three reports from the National Research Council (NRC) Committee to Review the U.S. Army Corps of Engineers Restructured Upper Mississippi River-Illinois Waterway (UMR-IWW) Feasibility Study. The committee’s initial report (NRC, 2004a) was primarily a review of a Corps of Engineers interim report (USACE, 2002). The following report represents a review of an updated and revised version of a study that was released by the Corps of Engineers in April, 2004 (USACE, 2004). It also contains comments on supplemental reports and other inputs, namely reports from the Sparks Companies, Inc. (of Memphis, Tenn.) on U.S. grain export forecasts (Sparks Companies, 2002) and from the U.S. Department of Transportation’s Volpe National Transportation Systems Center (of Cambridge, Mass.) on nonstructural measures for waterway traffic management (Dyer et al., 2003), and the Corps of Engineers’ Tow Cost Model. A third report from this committee is due in 2005. This committee’s statement of task was to review key issues, data, assumptions, and areas of controversy within the feasibility study, which includes economic, environmental, engineering, and plan formulation considerations (the full statement of task is listed in Chapter 1). This executive summary addresses the topics of integrated river system planning, ecosystem restoration, managing waterway congestion, forecasting river traffic levels, and the Tow Cost Model. It concludes with a brief summary of the report’s key points and recommendations.

The Upper Mississippi River-Illinois Waterway system is a natural resource of enormous value to the nation. As the largest riverine system in the United States and the third-largest drainage basin in the world, it is an

ecological resource of global significance. It provides an array of services to tens of millions of residents and visitors, including drinking water supply, boating, fishing (commercial and recreational), hunting, trapping, tourism, and commercial navigation. Communities, farmers, and tourists all use the river system’s floodplains, which provide sites for settlement, rich soils, and abundant recreational opportunities. These floodplains also experience occasional floods, which entail both costs and benefits. The same flood that threatens levees that protect farm land or communities also increases spawning habitat for fish on un-leveed floodplains. Historical modifications of the river to enhance commercial navigation, and land use changes across the watershed resulting from agriculture, forestry, and urbanization, have fundamentally altered hydrology and habitat. Similarly, actions aimed at ecosystem restoration may affect commercial navigation and other sectors. Sound river management decisions should consider the many values the river system provides, as well as interactions between users. The challenge is to find a balanced program that promotes individual uses while increasing the aggregate value of the Upper Mississippi River and Illinois Waterway to the nation. To its credit, the Corps has creatively expanded its feasibility study over time in an effort to broadly consider these interdependent issues, while also recognizing that the agency has neither the authorities nor the resources to plan and manage all of these issues across the entire UMR-IWW. Nevertheless, many of these uses and sectors are inseparable, and changes or impacts in one sector often affect other sectors. Declines in the value of any of these sectors are of national-level concern. A comprehensive feasibility study will consider these types of interrelations to help manage the system accordingly, and the Corps is correctly attempting to consider some of these relations between sectors.

The Corps of Engineers operates under a large body of authorities, legislative acts, and congressional committee language. This body of directives has accumulated over time without the benefit of any overall strategy or framework. Consequently, new authorities or legislation may be inconsistent with existing directives to the Corps. In instances in which guiding legislation or authorities are contradictory or unclear, the Corps is placed in the position of having to choose which authority or act will be given precedence. Within the context of the UMR-IWW feasibility study and the operations of the navigation system, the Corps considers the 1930 Rivers and Harbors Act that authorized the 9-foot channel project for the Upper Mississippi River as the primary management authority. Although a legal analysis of this situation was beyond the scope of this report, the current situation clearly poses ambiguities for the Corps and confounds the agency’s ability to manage the system in a way that maximizes its value to the nation. Although legislation subsequent to the 1930 Rivers and Harbors Act (e.g.,

the Upper Mississippi River Management Act of 1986) has proposed broader objectives for river management, the 1930 authorization is assumed by the Corps and some others to be the overriding authority in managing the resources of the UMR-IWW system. This single-purpose authority represents a barrier to the Corps and other agencies in adequately integrating other system values and uses, especially restoring river ecology, in UMR-IWW management decisions. The Corps and other UMR-IWW stakeholders recognize the limitations that this single-purpose authority imposes on trying to manage the system’s multiple resources. Consequently, in its feasibility study the Corps is seeking to add ecosystem restoration as an authorized UMR-IWW project purpose. This request is appropriate because it would enable the Corps to more holistically manage UMR-IWW resources in a way that better reflects today’s multiple users and diverse values (fulfillment of this request could also help clarify authorization priorities). Yet the request in itself may be insufficient for managing resources and issues that go beyond navigation and ecology, namely floods.

The Corps’ efforts to seek broader authority for planning and implementing projects on the UMR-IWW system are appropriate. The Corps should request a multiple-purpose planning and operations authority for the UMR-IWW, which would permit the agency to address flood management, navigation, and ecosystem restoration issues concurrently.

It should be emphasized that “restoration” of ecological structure, functions, and habitat in a system as large as the UMR-IWW is essentially unprecedented, and there is no perfect blueprint for restoration efforts on the UMR-IWW. Broad principles of large floodplain river science and management, as well as large river management experiences in the United States and around the world, may hold useful lessons for managing UMR-IWW resources (Europe’s Rhine and Danube Rivers, for example, are sites of ongoing restoration activities). It is also necessary to understand the constraints that the existing navigation system places on ecosystem restoration, and to explore linkages between the two purposes in a search for mutually beneficial actions. Although there are opportunities for improvement in the study’s ecosystem restoration component, the Corps has made good progress in this essentially unprecedented realm. Implementation of ecosystem restoration measures in the UMR-IWW will take place in a dynamic setting, with environmental and social changes that cannot be perfectly predicted, which suggests the necessity to manage adaptively. Part of “adap-

tive management” entails monitoring the impacts of management actions and using this information in future decisions.

The Corps has made progress by broadening the feasibility study to include a wide range of ecological restoration options, which entail hundreds of possible “projects” that could be implemented within the river-floodplain ecosystem. Although all of these projects conceivably have merit and may contribute to restoration goals, the prioritization, constraints, and prospects for success of these individual projects could be better explained; at present, it is difficult to understand which projects are the most promising and which should be implemented first. Although a clearly defined path forward for “restoring” river ecology in the UMR-IWW does not exist, the feasibility study’s approach to restoration could be clarified and its scientific basis strengthened by prioritizing actions that focus on restoring fluvial processes. Recovering some degree of these processes may at times affect structures, property, and human livelihoods; infringements upon these factors as part of ecological restoration must therefore be carefully considered. Linking proposed actions to larger, science-based themes and guiding principles for the ecosystem restoration component of the feasibility study, while representing implementation challenges of their own, would provide (1) a clear science-based framework for the feasibility study, (2) quantifiable metrics (e.g., flows, pool elevations, and acres inundated by floodwaters) for assessing progress toward restoration, and (3) a system for prioritizing management actions that have the best prospects for restoring ecological functions and processes. In moving to strengthen the scientific dimensions of the feasibility study and to explicitly incorporate adaptive management principles, the Corps should make full use of the federal-state Environmental Management Program (EMP). Established in 1986, EMP participants include the Corps, the U.S. Fish and Wildlife Service, the U.S. Geological Survey, and the five Upper Mississippi River basin states. The EMP was established to promote ecosystem restoration projects on the Upper Mississippi River, and to promote data gathering and ecological research. The EMP represents one of the nation’s best examples of a large river ecosystem monitoring program. Data collected and reports issued by the EMP should be an important part of the scientific basis for UMR-IWW management.

Since UMR-IWW ecology has important linkages with other sectors such as commercial navigation, recreational boating, and floodplain development, some future restoration measures may entail impacts on these other sectors. The menu of possible restoration measures is large. Some of those measures may entail only minimal disruptions to other sectors. Larger-scale measures, however, may well have noticeable impacts on recreation and commercial navigation. It is not clear when or where larger-

scale ecosystem measures might be enacted, and decisions regarding such actions will include input and opinion from the Corps and other federal agencies, state resource agencies, stakeholder groups, and the public. Yet because many, if not most, important future UMR-IWW management actions will entail trade-offs between sectors, information about how various sectors might be affected (perhaps including preliminary quantification of effects) is essential to good decision making. The feasibility study would be improved by a more prominent acknowledgment of sectoral trade-offs within the system, as well as discussions of the implications of future management decisions for trade-offs between sectors.

The ecological dimensions of the study could be strengthened by focusing efforts on restoring system-level hydrology and by broadening efforts to reestablish connectivity between the floodplain and river channel (or increasing the number of acres that can receive floodwaters during high flows) in areas where these connections have been disrupted by flood management projects and where there is support for alternative approaches (including willing sellers of leveed lands). The feasibility study should also more explicitly acknowledge and explain the interconnections between different users within the UMR-IWW and explore some of the key trade-offs that are likely to be part of future UMR-IWW management decisions. The feasibility study should be based clearly and explicitly on adaptive management principles, which rest upon both strong stakeholder collaboration and adequately funded, sustained monitoring programs. Management actions and policies regarding UMR-IWW resources should clearly incorporate and build upon past and ongoing monitoring and science programs of the federal-state interagency Environmental Management Program.

The current system for managing congestion at locks on the UMR-IWW is largely a “first-come, first-served” system. The issue of alternative, nonstructural systems for managing waterway traffic has been a topic of considerable discussion, with some groups (including this committee in its first report and the “Phase I” committee) recommending implementation of more efficient management systems. To respond to calls for more efficient waterway traffic management systems, the Corps has considered a variety of nonstructural measures for managing waterway traffic congestion. Several promising traffic management strategies have not been adequately considered or evaluated in connection with the feasibility study, however, including combi-

nations of real-time scheduling and lockage priority rules (as presented in several different papers, some of which were published by the Corps; see, for example, USACE, 1999), “slots trading” proposals, or congestion fees. Although there are signs that these and other potentially useful nonstructural measures are being considered or are under development, there is no guarantee that such measures will be vigorously pursued and implemented in the future. The near lack of any analysis of the viability of nonstructural measures for managing waterway traffic represents a considerable analytical gap within the feasibility study, because it is not clear how the benefits of lock extensions can be reliably estimated without first managing waterway traffic more efficiently within the existing system.

More efficient nonstructural measures for managing UMR-IWW traffic are needed regardless of whether lock extensions on the UMR-IWW system are constructed; if extensions are not constructed (or if it is decided to delay construction), more efficient management of the system will be necessary to improve operational efficiencies. For example, as part of the president’s fiscal year 2004 budget background information, the executive-level Office of Management and Budget recommended that for inland waterways navigation projects, “The Corps should make greater efforts to reduce traffic congestion through scheduling and other demand-management approaches.” If extensions are constructed, more efficient nonstructural measures will be necessary to alleviate the reduced shipping capacity of the system during construction. Without such nonstructural measures, construction may result in more temporary congestion than is necessary. In either case, the Corps and others may be missing an opportunity to immediately lower transportation delays and costs. Like efforts to “restore” river ecology, creating a useful nonstructural traffic management system on the UMR-IWW is not a simple matter. Progress on this front, however, should be guided by the same adaptive management principles being considered for ecosystem restoration: build on existing experience, knowledge, and approaches (for example, the U.S. Coast Guard has developed a “vessel traffic service” [VTS] system that may have possible applications to the UMR-IWW), implement thoughtful and deliberate initial steps (e.g., a small pilot project(s) could be useful), monitor results carefully, and make appropriate adjustments and changes through time.

Implementing some nonstructural measures for managing waterway congestion could decrease congestion, reduce shipping costs, and use the existing waterway more efficiently. Because the costs of implementing nonstructural measures are low, and because some have positive net benefits, implementation of these measures should be of the highest priority. A comprehensive evaluation of UMR-IWW waterway traffic management alternatives will identify and thoroughly evaluate all plau-

sible measures. The failure to consider and evaluate the prospects of all potentially beneficial nonstructural measures for better managing waterway traffic undermines the feasibility study’s conclusions and recommendations regarding proposed structural improvements.

Forecasts of UMR-IWW traffic levels figure prominently in the consideration of lock extensions. Forecasts of long-term increases in towboat traffic on the waterway would strengthen the case for extending locks from 600 to 1200 feet, while forecasts of stable or declining levels of exports provide less support for extensions (and likely imply that extensions are not economically justified). The creation and application of credible forecasts represent key analytical challenges within the feasibility study, because locks have a design life of several decades and the accuracy of waterway traffic forecasts diminishes considerably after 5 to 10 years.

Current commercial shipments on the UMR-IWW consist of roughly 50 percent grain (corn, soybeans, and wheat) and roughly 50 percent other goods (construction materials, petroleum products, coal and coke, chemicals, and fertilizers). A set of forecasts of U.S. grain export futures was produced by Sparks Companies, Inc., of Memphis, Tennessee for use in the Corps study. Consistent with advice from the NRC Phase I committee’s 2001 report, a range of forecasts (rather than a single forecast) was produced. Five forecasts were presented in the 2002 Sparks Companies report. One of those forecasts reflected a scenario of a slight decline in exports over time. The other four forecasts presented increasing levels of U.S. grain exports over time. Today, none of these forecasts can, of course, be either fully verified or discounted. One of these five forecasts may eventually turn out to have been correct, U.S. grain exports may exceed even the greatest forecast increase, or exports may decline even farther than the “steady-to-low-decline” scenario. A range of global supply and demand variables, or “drivers,” will affect future levels of U.S. grain exports, including income levels and consumer preferences in China, grain production policies and output in Brazil, changes in water use and supply in grain-producing regions, and prospects for local, U.S. domestic grain processing. Credible forecasts of U.S. grain export levels will consider trends in these drivers. Notable trends in one or more of these drivers may strongly influence a grain export forecast.

The presence of uncertainty does not mean that all futures are equally likely. The future flows from the past, and in the absence of changes in important “drivers” that interrupt this flow, the near future will continue to

be much like the recent past. Sound forecasts should begin with a careful examination of the most important drivers that have shaped past trends. They should then examine how these drivers are expected to change, and whether new drivers are likely to be introduced. Widely accepted data show that U.S. grain exports since 1980 have been relatively steady. Although future changes in global grain demand and supply drivers are likely, there are no overwhelming regional or global trends that clearly portend a marked departure from a 20-year trend of steady U.S. grain export levels. U.S. grain export levels may indeed increase in the future (as suggested in the scenarios in the Sparks Companies report and in those from the U.S. Department of Agriculture), but there are also several plausible changes in global drivers that could lead to declining exports. For example, population growth and declining water tables in China could contribute to increasing demand for U.S. grain to feed to livestock in China. On the other hand, increases in income that affect consumer preferences in China could support increased demand for U.S. meat exports to China, which would replace exports of feed grain. Increases in the amount of irrigated cropland in China, and attendant increases in Chinese grain production, could also dampen demand for U.S. grain. Similarly, changes in U.S. agricultural subsidies and policies could influence grain exports in either direction. For example, some legislative incentives could encourage increased grain production; other incentives could encourage increased ethanol production, which would tend to reduce grain exports. Increases or decreases in export levels will be driven by changes in these types of factors that influence supply and demand in global and domestic grain markets. Absent these types of changes, future levels of U.S. grain exports, at least in the near term, are likely to resemble levels of the present and the recent past.

It should also be remembered that non-grain freight on the UMR-IWW constitutes about 50 percent of movements on the waterway, which introduces additional uncertainty regarding the movement of commercial goods. The non-grain barge traffic forecasts within the feasibility study are based on only one scenario for future movements of each of seven key non-grain commodities, as opposed to the multiple scenarios presented for grain shipments. Each non-grain commodity group forecast was developed separately based on models constructed in the mid-1990s. Traffic levels of nearly all of the seven non-grain commodity groups are forecast to increase about 20 percent over a 50-year period. If additional non-grain scenarios that included the possibility of steady or decreasing traffic levels were presented, the result would be a more complete and balanced set of barge traffic scenarios.

Non-grain shipments on the UMR-IWW are roughly half of total commodity shipments. A more complete set of scenarios of UMR-IWW waterway traffic would thus, in addition to considering the pos-

sibility of both increases and decreases in grain commodity shipments, consider possible increases and decreases in non-grain commodity shipments.

The Corps’ Tow Cost Model, or TCM, was originally developed by the Corps of Engineers Ohio River Division to model coal shipments along the Ohio River. The TCM has subsequently been used in the feasibility study to model changes in grain and other commodity shipments on the UMR-IWW in response to changes in shipping costs. The Tow Cost Model, however, has limited relevance and value within the UMR-IWW feasibility study. In particular, the TCM assumes that changes in shipping rates do not result in any traffic moving off the waterway to alternate modes until rates reach a specified level, a point at which all traffic then moves off the waterway. Because of this behavioral assumption, under specific circumstances the TCM can produce an “upper bound” on benefits of lock extensions. It is the case, however, that users such as grain shippers often have a variety of alternatives for responding to changes in shipping rates that include local processing of grain, shipping grain via an alternate mode, shipping grain via an alternate route, or shipping grain to a different destination port. Similar choices are available to shippers of non-grain commodities. These possibilities, however, are not considered in the TCM. For this reason, the TCM cannot be used to accurately estimate the benefits of reducing congestion. Instead, it estimates an approximate upper bound on economic benefits. The Tow Cost Model contains assumptions and functions that do not adequately reflect responses of shippers to changes in shipping costs. It therefore produces results that are of only marginal use in the feasibility study.

• An UMR-IWW feasibility study that integrates commercial navigation and ecological restoration plans represents an unprecedented analytical challenge, and the Corps of Engineers is to be credited for broadening the scope of this effort during its study. The Corps has especially made substantial progress on the study during the past three years. Despite these efforts, the study contains flaws serious enough to limit its credibility and value within the policymaking process.

• Despite statements to the contrary, the commercial navigation and ecosystem restoration components of the study have proceeded essentially on separate tracks. There is little evidence that prominent and important trade-offs and interactions between these two sectors and other users on the UMR-IWW were carefully considered.

• There has been more attention devoted to considering nonstructural measures in the April 2004 draft feasibility study than in previous versions. The study still lacks adequate analysis in this realm, however, and some promising nonstructural approaches for managing waterway traffic appear not to have been considered at all. The failure to fully consider nonstructural measures precludes any statement about the desirability of structural measures.

• The approach to estimating the benefits of lock extensions, using scenarios and upper and lower bounds on benefit estimates, is a reasonable way to deal with the uncertainties that attend the long lifetimes of inland waterway infrastructure, but the scenarios of future waterway traffic levels remain problematic. On further consideration, the largest concerns focus on the single forecast of non-grain commodities, and the scenarios reflect a bias in the direction of future increases in exports.

• The Tow Cost Model used to estimate benefits of lock extensions can provide no more than an approximate upper bound on benefits. In an attempt to deal with the TCM’s shortcomings, the Corps developed the “ESSENCE” model. This model incorporates a hypothetical demand curve to estimate benefits of proposed lock extensions, but the curve has no empirical foundation. The ESSENCE model is therefore incapable of producing any credible estimate of a lower bound of the benefits of lock extensions. Economic feasibility has therefore not been demonstrated for any of the navigation alternatives.

• Substantial effort has been devoted to designing an ecological restoration program, and the proposed restoration measures represent an impressive range and number of candidate actions. The assembly of these measures into restoration alternatives, however, is not adequately grounded in principles and theories of large river floodplain science and restoration. For example, evaluation of the candidate alternatives relies, in part, on metrics (e.g., “area affected”) that are poorly correlated with ecological outcomes. The ranking of alternatives is therefore unpersuasive.

• The Corps’ Preferred Plan, if carried out as described, provides for a program of incremental implementation, an excellent framework for comprehensive adaptive management. If the Corps is provided the resources—and if it commits to the needed data collection, improved modeling techniques, and evaluation—many of the flaws and omissions in its study can be corrected in the course of implementation by application of adaptive management principles.