The CAN-Do Approach

January 1, 2002

Users and the environment frequently place competing demands on water to meet a wide range of objectives. How water resources managers response to these demands is monitored by a complex mixture of allowable actions, goals and barriers that must properly evaluated and integrated in order to come up with a satisfactory mix of benefits. Our ability to control runoff is limited. We can build reservoirs to hold water for later use, control water use and construct flood control facilities. We can also change the way existing facilities are operated. But, we still can't stop (or start) the rain.

How do you operate water supply systems, such as dams, reservoirs, canals, pipelines and water treatment plants? How do you use the available water? What do you build? Even though the range of things to do is limited, the number of different combinations of these things is very large, and finding a combination that satisfies all users is often a daunting task.

HydroLogics, a water resources management consulting firm, has developed a process for conducting consensus building negotiations -- a process that, more often than not, delivers win-win solutions and benefits ranging into the hundreds of millions of dollars.

The HydroLogics process uses its OASIS simulation software -- with patented logic routines -- to facilitate computer-aided negotiation (CAN) sessions. This process, along with the software, have been or are being used to resolve water management issues in South Florida (Everglades), the Delaware Basin, the New York City Water Supply System, the Washington Metropolitan Area, the Roanoke River in North Carolina, the Kansas River, the Las Vegas Valley and many other places. It was a vital factor in the development of several new water institutions, including the Cooperative Operations Section of the Interstate Commission on the Potomac River Basin, the Kansas River Assurance District and the Southern Nevada Water Authority.

How It Works

A small, operations-based increased in the effectiveness of existing facilities can provide enormous benefits.

The CAN Process

The CAN process shares much with mediation, but the "mediators" often take a much more active role in determining the outcome than most are comfortable with, and do so with less clearly defined and agreed upon procedures. The CAN process starts by identifying performance measures (PMs), which reflect the parties' interests and can be used to compare alternative solutions. Most PMs are not textual, but are graphs, charts, pictures, checklists and other highly visual, well-defined representations of the multiple objectives. They allow parties to distinguish better from worse when comparing objectives and must be based on some reproducible evaluation method. The process proceeds by obtaining consensus on the data and methods to be used in the evaluation.

This is followed by a compilation of alternatives. Most often, operational and other non-structural alternatives are at least as important as physical alternatives. This is because the installed base of facilities is usually much larger than proposed new facilities. Thus, a small, operations-based increase in the effectiveness of existing facilities can provide enormous benefits.

The design of the actual analytical techniques is the next step in the process. OASIS software is a powerful water resources simulation engine for generating long- and short-run evaluations; it also serves to unify the evaluation of surface water, groundwater, water quality, fisheries management and other analytical tools through its standardized run-time interface for linking in "side" models. This interface lets the models run together, "in parallel," and to communicate back and forth at each time step.

OCL deals both with multiple operating constraints, which must be met, and multiple operating targets, which operators try to achieve and which they must "balance" when they all cannot be achieved simultaneously.

New York State CAN

The patented Operations Control Language (OCL) used in OASIS allows the user to describe operating policies to the underlying OASIS model. This is crucial to the success of the analytical portion of the process. The syntax and keywords of OCL are simple, mnemonic, and reflect the ways in which actual operators tend to describe operating policies. OCL deals with both multiple operating constraints, which must be met, and multiple operating targets, which operators try to achieve and which they must "balance" when they cannot all be achieved simultaneously. As a result, most people knowledgeable in water resources can easily read and understand OCL, and this makes the analysis much more transparent to users. In other words, OCL was designed for water people -- not computer programmers.

Models running in parallel are linked through OCL, which allows users to write rules that depend on the parameters simulated in side models (e.g., rules based on water temperature or quality, rules based on provided habitat for migratory species at the time of arrival, rules based on highly differentiated water demands, etc.). OCL also allows the users to set parameters inside the side models (e.g., the rate and location of groundwater pumping for the current time step). Most importantly for the CAN process, OASIS with OCL makes it possible to modify alternatives and produce new results quickly.

During a CAN session, the HydroLogics staff plays a very proactive role. They serve as a human interface between the parties to the dispute and the computer, and serves an engineering design function, as well. Guidance for maintaining neutrality while being proactive is grounded in multi-objective theory, and recognizes that values come into play in the choice of objectives and in the allocation of benefits. Therefore, the HydroLocis staff are very careful not to engage in making choices in either of these areas. The staff objective in engineering design is to create alternatives where each objective is fulfilled as much as possible without reducing the achievement of other objectives. Such alternatives are called "non-inferior" in multi-objective parlance. The choice of which of the non-inferior alternatives to pick is entirely the province of the disputing parties.

OASIS with OCL

The OASIS software that makes CAN possible is not some fully generalized simulation package, but is specifically designed to address water-resources related problems. Characterized by ease and flexibility of use, OASIS "knows" the basics of how water supply systems work and automatically handles such things as: conservation of mass, reservoir characteristics, channel capacities and minimum flows, annual patterns, time series inputs, etc. Standard input tables are provided for such items, and this speeds up the development of analytical tools.

Win-Win Solutions

The CAN process is very useful for resolving many complex water resources disputes. Because of its early and thorough focus on PMs and because it resolves data and scientific methods disputes prior to any evaluation of alternatives, it creates a common basis for evaluating and illuminating the pros and cons of alternatives. This is extremely important in developing the consensus required for non-judicial resolution of disputes.

How It Works

A Computer Aided Negotiation session proceeds as follows:

1) The HydroLogics staff begins with a presentation of the evaluations of current operations and facilities (the base case) and three to four other alternatives which improve system benefits in terms of the performance measures. It is generally possible to produce alternatives that are win-win in terms of the stated performance measures compared to the base run.

2) Problems with the runs in terms of benefits, or equity or details of the rules and facilities are noted, along with specific suggestions for modifying alternatives. The participants agree on the next alternative or a small set of new alternatives to be evaluated.

3) The staff modifies the input data and OCL using the OASIS graphical user interface, and makes the runs while the participants take a short break. This usually takes from 15 to 30 minutes. OASIS outputs graphs, charts, and other performance measures.

4) Steps 2 and 3 are repeated until a consensus solution is reached.

It is usually possible to go through steps 2 and 3 from four to five times in a single day. After the session, the staff summarizes the results and attempts to find even better solutions based on what occurred in the CAN session.

New York State CAN

In 1954, the U.S. Supreme Court issued a decree that allocated the waters of the Delaware River in New York City (NYC) and the Basin States (New York, Pennsylvania, New Jersey, and Delaware). The Delaware River Compact empowers the Delaware River Basin Commission (DRBC) to modify the decree given the unanimous consent of all of the parties. Proposals to modify operations are generally considered to modify the decree, and are a frequent source of discussion.

In this case, the New York State Department of Environmental Protection (NYSDEP) desired to change the pattern of minimum releases from the NYC reservoirs to enhance the trout fisheries in northern Delaware. The OASIS model being developed for the DRBC provided an analytical framework for focusing the negotiations.

Initially, evaluations were made in order to compare existing operations and policies with the NYS proposal. These were presented in a CAN session. Their proposal impacted both the reliability and quality of the NYC supply, and was not acceptable to NYC. The proposal also seemed to have a large number of years in which too much water was committed too early, causing potentially severe late-summer impacts on the fisheries when the water dedicated to temperature control ran out. Comments on the results indicated the nature of the problem, and potential operational changes to overcome the problems were discussed.

Prior to convening a second CAN session, HydroLogics developed several new operational alternatives which reduced impacts on the NYC water supply to very low levels and also improved the performance for the fishery. These were presented to the parties at the second CAN session. The parties then proceeded to modify one of the alternatives to produce an alternative acceptable to all.