Water Modelling

In this Section:


Introduction

The use of computers has provided the opportunity to better understand and assess our water resources through comprehensive numerical model simulations and testing of various schemes or options. The numerical models, a combination of equations that represent physical structures and their hydraulic impact upon lake and streamflow, allows the user to assess the hydraulic conditions in the basin and thus, establish a better understanding of human impacts upon a natural or modified river and lake system.

In terms of regulation modelling, the modeller can incorporate various user interests, as well as historical uses, to generate operating scenarios to verify the variations, alternatives of interest to the basin community, as well as the physical environment, that has been changed. Assessments of operating policy changes, impacts of floods, and changes in water quality are just a few examples where numerical models are used. Physically based models determine the flow and level changes that are currently being employed to determine the impact of man-made changes upon the river and its active biological community.

Newer models, such as those used to predict ice movement, are being developed to address the northern river problems such as ice damage to river banks and bridges, and increased impacts of flooding. Given that the river is an ever-changing biological entity, water quality under ice conditions has yet to be explored and addressed through numerical means.

Irrigation and water use conflict models are now becoming more popular to address the ever changing needs of basin managers. More complex riverine situations such as deltas and multi-channel irrigation systems are now becoming better understood by the application and further development of numerical models that can address larger systems problems.

The ability to study and project multi-sectoral water uses in relation to social, economic and other considerations and their impacts on the water balance of a basin is an important aspect of water modelling. The importance is further magnified when such ability can be projected onto global warming induced climate change scenarios and/or interjurisdictional basin studies to allow the formulation of interjurisdictional or international water apportionment agreements.


Water Management Models: Types and Applications

Simulation

Simulation models compute flows, levels, water quality, and other parameters based upon physical laws and conceptual relationships. The models are calibrated by comparing computed results to observed values with subsequent adjustments to mathematical coefficients to attain closer agreement. The models are verified by comparing historical events not used in the calibration to computed results. Simulation models are widely used for both planning and operational purposes.

Hydro-Configuration Modelling System (HCMS) Model: A generalized planning model that can be used for large reservoir/channel networks. The solution is based upon the IBM linear programming optimization routine DOSPB of the package SL-MATH. The model solves time steps individually (rather than simultaneously, like the models REGUSE and ORRMS).

Application: Planning studies for major river systems – A linear programming model of the river system is generated to represent the physical characteristics of individual reservoirs, control structures, and river channels; linkage characteristics describing the river/reservoir/hydro-generator network; dynamic constraints relating to the permissible operation of the system, and operating controls or targets. Variables are reservoir storage/elevation, reservoir discharges, and violations of specified bounds (computed inflow file required).

Steady-state and dynamic (transient) models

Steady-state models assume constant discharges (flow, effluent), which very seldom occur in nature. However, the assumption of steady-state conditions is usually on the conservative side, especially in flood plain hydrotechnical studies. Advances in computer hardware and software are increasing the popularity of the more physically based dynamic (transient) models, which compute variable values as a function of time.

ONE-D Model: One-dimensional hydrodynamic implicit finite-difference simulation model with optional water quality routine.

Application: Hydrotechnical planning studies and operational level and flow forecasting – Simple single-channel to large multi-channel networks and estuaries. Water quality modelling in river and estuarine studies.

RIVICE Model: A transient (dynamic) simulation model in which major ice phenomena and processes are considered. The model is currently under development with financial support from 13 agencies (2 from the U.S. and 1 from Sweden). Another 12 agencies are providing technical assistance and test data.

Application: Hydraulic simulation – Ice cover formation and ablation, frazil ice formation, anchor ice, ice transport, hanging dams, breakup, and jams. The ONE-D model has been selected as the hydrodynamic module. The model is to be used for hydrotechnical studies relating to flood damage reduction and environmental impact of proposed structural and flow regulation changes in river networks. The model will be employed by hydroelectric utilities for operations, and is also likely to be used widely for adjusting flow measurements taken under winter conditions.

Operational research (OR) models

OR models have decision-making capabilities within the simulation, subject to user-specified constraints on the variable values that are being computed. The models usually try to optimize some parameter such as maximizing net benefits or minimizing losses. OR models are usually much more complex than simulation models but can arrive at a much more direct solution than the trial and error approach used in simulation models.

ORRMS/MORRO Model: A linear-separable programming operational model permitting an integrated operation of hydroelectric stations, reservoirs and channels. Model is dependent upon the IBM software package MPSX.

Application: For optimal flow regulation the model can optimize the operation of a complex reservoir/channel network, taking into account flood damage and energy production while respecting constraints on flows and levels imposed by other interests such as recreation, navigation, and log driving. The model was developed mainly to facilitate the operation of major river systems, however, it can be used for planning purposes such as determining the impact of new reservoirs and assessing the effect of new constraints. The model is generalized and can be applied directly, or with only minor modifications, to other river systems (requires selection of an appropriate inflow/runoff model to provide input in operational mode).

Expert systems (ES)

ES is a special field of artificial intelligence, which is used in water management for decision-making. ES uses a collection of facts, rules of thumb, and other knowledge to help make inferences on how to deal with the water management problem under consideration. Expert systems differ substantially from conventional computer programs in that their goals may have no mathematical solution, and they must make inferences based on incomplete or uncertain information. They are called expert systems because they address problems normally thought to require human specialists for solution. Their success lies in their ability to analyze large amounts of information according to pre-established rules resembling the reasoning of a human expert or group of experts.

Water Use Analysis Model (WUAM): This is a water resources planning model which integrates water supply and demand considerations into a basin water balance and overall management framework. WUAM is a highly flexible, interactive micro-computer simulation model which places particular emphases on water use. Using socio-economic, hydrologic, water use and other data, and information/assumptions about future growth, changes in technology, water use practice trends and policy assumptions regarding water pricing, the user can obtain projections of multi-sectoral water uses, compare the projected water uses against available supplies, and derive, among numerous other details, statistics about the severity and frequency of water shortages, if any. Other features of the model include the capability to consider water diversions and interjurisdictional water apportionment, to analyse the impacts of water price on water use, to model reservoir operations, to account for water use priorities, and to analyse water rationing, when available water supplies are approached or exceeded.

Applications:

Water Supply Constraints to Economic Development – Water demand projection from a multi-sectoral viewpoint for water availability/constraints investigations for economic development.

Water Conservation Studies – Water demand management tool for studying the impacts of various water conservation measures on such future demands; e.g., water water pricing, metering and recycling, etc.

Climatic Change Impact – A potential tool for examining climatic change impacts on future water demands and supply-use balance.

Interjurisdictional Basin Studies – WUAM's ability to consider flow apportionment at interjurisdictional boundaries makes it suitable for international and interprovincial river basin studies. It can be used to quantify the impacts of developments or growth in various sectors on international and interjurisdictional water apportionment agreements.

Hybrid expert systems

Hybrid expert systems are now appearing in water management modelling. These models are characterized by a heuristic database (intuitive data and information garnered from experts) and a combination of mathematically and physically based simulation and operational research techniques.

REGUSE Model: A generalized model using a network flow optimization algorithm and a heuristic database for both planning and operational modelling of flow regulation and multiple-use multi-reservoir/channel networks. Time horizons for which simultaneous solutions are obtained can be chained for planning study simulations. Unit-time periods can vary from one day to a month. The use of daily data and extended time horizons permits inclusion of channel routing in the solution process.

Application: Flow regulation and water use planning studies (basic input requires computed inflow files at model nodes, stage/storage/discharge relationships and heuristic data consisting of rule curves and penalty coefficients for violation of bounds). Basin-wide operational flow regulation studies (requires selection of an appropriate inflow/runoff model for input in an operational mode).


Mathematical Modelling: Project Descriptions

Mackenzie Delta-Beaufort Sea water quantity and quality model studies

Water in the basin offers habitat for fish and provides nutrients for the food chain which supports fish and wildlife. Coal, gas, oil, tar sands and minerals require large amount of water for their development. Navigation and hydroelectric power production are other important water uses, as is community needs in the North. Development of any aspect of water affects other uses, forexample, by altering winter or summer water levels, changing water quality, sediment load and water temperatures, contaminating groundwater, etc. The One-Dimensional Hydrodynamic model (ONE-D) is being applied to simulate the downstream hydrologic effects of water resource projects so that better understanding can be gained of how such projects will impact the river (and particularly the Delta) and how potential damaging effects of such projects can be mitigated.

Peace-Athabasca Delta environmental model studies

For centuries, the Peace-Athabasca Delta experienced recurring summer flooding followed by gradual recession of waters during the fall and winter. Hydro development on the upper Peace River in the late 1960s altered the hydraulic flow regime affecting the Delta, and the plant and animal life that had become dependent upon frequent flooding was threatened. Permanent weirs in two channels draining the delta were put in place in 1976 to reduce extreme low levels. The ONE-D model has been applied in the study the Delta's complex hydraulic network and to determine the effects of flow regulation, both upstream and downstream of the Delta, to evaluate the mitigating effects of the weirs constructed in 1976, and to assess the effect of 1) a proposed control structure at the outlet of Lake Mamawi, and 2) a proposed dam downstream on the Slave River. Simulations over a 22-year period were conducted in this study as opposed to the normally short simulation of several days required for flood studies.

In the current Peace-Athabasca Delta Ecosystem Management Plan Development Studies, the model is being applied as a major tool to simulate various flood scenarios caused by ice jams as part of the development of strategies to restore the lowering delta water level.

Fraser River water quantity and sediment transport model studies

The Environment Canada ONE-D model is employed operationally on the lower Fraser River for the simulation of flow in a tidal-affected regions where flow reversals temporarily occur in several reaches. The model computes instantaneous, hourly and daily discharges. These discharges are subsequently used in the computations of suspended sediment loading calculations.

Qu'appelle Valley water management and planning model studies

In order to assist the Qu'Appelle Valley basin managers in the real-time operational control of river flow and storage waters and in carrying out various types of basin planning studies, the model REGUSE has been developed. The basis of this water balance model is the simultaneous consideration of all hydrologic events (runoff/tributary inflows and evaporation loss), hydraulic characteristics (stage/storage data, routing, etc.) and economic water use data for a specified time period. Real-time applications using either short or long time periods can be run with hydrological forecast and climatological data. The use of extended time periods permits channel routing to be included in the solution process. The model automatically generates a complete flow network for the basin. The "out-of-kilter" optimization algorithm is then used to compute the solution considering a network flow approach.

Red River flood forecasting model studies

Frequent flooding in the Red River Valley south of Winnipeg (Manitoba) led the provincial and federal governments to undertake joint flood protection forecasting studies. One part of the project is to produce a physically based mathematical routing model capable of predicting flows and levels from Emerson to Winnipeg (145 km along the channel). The routing model can take into account: mild surface water slopes; channel and off-channel storages; dyke, bridge and road overtopping; multi-channel flow; uniformly distributed or point lateral inflows resulting from time-variable rainfall and snowmelt; changing river ice cover. The model is also being employed for studying the effects of dyke and road construction along the Red River, south of Winnipeg.

Great Lakes regulation model studies

The SYSNET model is a specially designed optimal solution seeking simulation for regulation of the entire Great Lakes system in which Lake Nipigon, Ogoki-Long Lake, Chicago, and Welland diversions are included in the model configuration. There has been a demonstrated need to consider the Great Lakes and its diversions as a single integrated system. The objective function (optimization goal) of the model essentially consists of keeping water upstream for future needs as allowed within the specified flow and level constraints for each monthly period. The network flow technique (optimizer) used in this model is based upon what is generally referred to as the "out-of-kilter" algorithm.

St. Lawrence River and region of Montreal model studies

The Environment Canada ONE-D model was also applied to a flood study in the Montreal region. The problem consisted of determining water level elevations produced by highly regulated flows of the St. Lawrence River meeting the partially regulated flows of the Ottawa River in the complex hydraulic region of Montreal. Since there is limited storage capacity in the Montreal area, a steady-state analysis of separate flood frequency flows was performed with the model to determine flood levels for the design of dykes and flood line contours.

Upper St. John River water quality model studies

The ONE-D model has been used to simulate the flow and water quality of the St. John River and to determine its responses to eight water quality variables under high and low flow conditions. The reach from the upstream end to Grand Falls is included in the study and involves participation from the Atlantic Region and the Maine state Department of Environmental Protection.

St. Croix River water quality and quantity model studies

Two water quality models were applied to the St. Croix River for a comparison study of the river's response to different pollution loadings under various flow conditions. Simulations of dissolved oxygen (DO) and BOD were carried out in collaboration with the Atlantic Region in 1983 and 1984.

Truro flood damage reduction hydraulic study

A flood study conducted with the Environment Canada ONE-D model was carried out in the Truro region of Nova Scotia. The North and Salmon rivers, during storm events, carry flood discharges down steep gradients to their confluence in the Town of Truro. There, flood flows and high tidal cycles on the Salmon River estuary can cause severe floods in the town. Dykes along the river with aboiteaux structures cannot entirely accommodate the 100-year and 20-year storm events. Consequently, simulations of these flooding events must be jointly considered in flood flows and tidal cycles; flows through aboiteaux and over dykes; and flows that leave the main channel as it is overtopped and move downstream in defined flood channels eventually to rejoin the regular river.

Ottawa River Basin flood and hydro-electric regulation model studies

A generalized modelling package referred to as HCMS (Hydro Configuration Modelling System) has been developed for studying multi-reservoir river systems. The Package is particularly applicable to such basins as the Ottawa in which there is existing hydroelectric generating capacity and/or a potential for further development. The package is used to evaluate proposals relating to: increased reservoir storage devoted to flood protection; new reservoirs; new installed turbine capacity; environmental requirements in terms of constrained reservoir levels or minimum flows; diversions of flows; and increased channel discharge capacity. By comparing the results of different simulated operation policies and configurations with each other and with historical performance, the planning team can more readily assess the combined impact on flood protection, energy production and environmental concerns.

Water use analysis of the Exploits River Basin, Newfoundland

The appropriate planning of water resources within a river basin requires, among other things, information on, and understanding of, the available water supplies and the present and future uses and demands for water. In the province of Newfoundland, the single purpose use of water resources has been permitted in the past, without regard to other users. These exclusive water rights have resulted in conflicts with other water uses in the basin and were recognized as an obstacle to the optimum use of the available water resources.

The Water Use Analyses Model (WUAM) was used to assess the water supply and water use conflicts in the Exploits River basin, in order to provide the province with the necessary information for the rational management of the water resources in the basin, with due consideration of all water uses. The Exploits River basin has a drainage area of approximately 11 000 km2 and is the largest basin on the island of Newfoundland. The basin is well forested with forestry/pulp and paper being the largest single industry and the main consumptive use of water. The primary water demands are instream uses for hydro power generation, fisheries, recreation, and use as a receiving water for domestic and industrial waste water effluents. The basin also has five hydropower generation facilities.

The WUAM model was calibrated for existing conditions. The calibrated model was then used to assess the impacts and water use conflicts for a number of possible development scenarios. Results obtained from the model provided the rationale for future developments in the basin, in particular for hydropower generation, which would not conflict with the other uses.

Date modified: