RED RIVER RISING
By Mark Bennett, M.P.A., C.E.M.
The City of Winnipeg, capital of Manitoba, with a population of about 650,000, is at the geographical centre of North America. Located about 829 miles north of Kansas City, Winnipeg is at the confluence of the Assiniboine and Red Rivers.
The Red River Valley has a long and storied history of flood events. The flow rates of the River are notoriously variable with all-time minimum flows of about 13 cubic feet per second (cfs) and an all time maximum flow of over 200,000 cfs. Each spring brings with it a threat of serious flooding. With head waters in the United States each year’s problems are first manifest there (North Dakota) and the crest gradually moves north into Canada, eventually through Winnipeg (see map 1 on page 54). This south to north flow allows for a further complication in as much as ice cover is generally present on the River at the commencement of spring flooding. Furthermore, points down stream are usually frozen longer than up-stream points. This significantly compounds the likelihood and severity of ice jamming and blockages as a complicating factor.
In this article I will highlight some of the key events and lessons learned in the flood of ‘97 or the “Flood of the Century”. A brief explanation of the factors which contribute to flooding will be given. Mention will also be made of how available forecasts prompted appropriate warnings. The City of Winnipeg benefits greatly from one of the most effective disaster mitigation projects in the world. The article will touch on the Red River Floodway and its key role in this disaster. The impact, or when the crest (highest river levels) reached Winnipeg, was the culmination of all the preparedness activities. That was when all defences met their sternest test. This is also the time when the most number of people were evacuated from their homes.
There are five factors which contribute to flooding in the Red River valley. In looking at the anatomy of a flood the first variable to come into play is soil moisture levels (factor 1) going into winter and freeze up. As the valley itself is actually an ancient lake bottom heavy clay laden soils predominate. The reduced ability of these soils to absorb surface water is the second factor. Thirdly, snow accumulation (factor 3) over the winter can signal increasing probabilities of serious flooding. The winter of 1996-97 was very severe throughout the watershed. Some areas around Grand Forks, North Dakota reported three times normal snow depth into late winter early spring. The rate at which the snow melts (factor 4) largely determines how rapidly surface waters reach the river and tributaries and in what volumes. This year the thaw was late and relatively quick. And finally the fifth factor is coincidental precipitation (rain or snow). Significant amounts can add substantially to an already serious situation. This year just as the event was beginning during the first weekend in April almost the entire valley experienced the worst blizzard in over 30 years. The City of Winnipeg received one and a half feet of snow and reports were that the situation was even worse throughout eastern North Dakota.
Tragically, for the City of Grand Forks, this blizzard probably could not have come at a worse time. For Winnipeg, being downstream, with later onset flooding, we had the opportunity to recover from the blizzard and adjust our flood fighting plans accordingly.
Each year, starting in February, the noted flood factors are assessed and forecasts are generated. Obviously at this early stage the only variables with some clear definition are the soil moisture condition and the amount of snow cover, which is assessed as a water equivalent. In February 1997, the preliminary indications were that, even with normal weather accounting for both snow melt and coincident precipitation, the situation promised to be very challenging. If adverse conditions occurred it was predicted that the worst flood of the century was likely.
Typically the forecasts indicate the ranges within which the river is expected to rise. The upper end of the range essentially constituting a worst case scenario. From the outset the City of Winnipeg resolved to plan for the worst, but hoped for the best. The City’s plan for fighting the flood was quite simple. Properties at risk, given the predicted levels, were identified and surveyed. Each property would require the construction of a sand bag dike (levee) to protect it two feet (freeboard) above the forecasted level. The number of sand bags required on each property was determined based on the required length and height of diking. Implementing the plan was much more difficult. The first forecast called for a 22.5 foot rise in the river, jeopardizing 256 homes and requiring 2.6 million sand bags for dike construction.
Over the course of the next nine weeks the forecast was changed periodically due to changing conditions. The revision after the blizzard resulted in a two foot increase in the forecast. In the final analysis close to 700 homes were at risk requiring over six million sandbags to establish and maintain protection.
The City of Winnipeg is the main beneficiary of a major flood control system which includes three geographically separate projects. They are the Shellmouth Reservoir, the Portage Diversion (both of which are on the Assiniboine River) and the Red River Floodway.
The Floodway is one of the largest excavation projects in Canadian history. Over 100 million cubic yards of earth (40% as much as the Panama Canal) were removed to construct a diversion channel 29½ miles long. The average depth of the channel is 30 feet and it’s width ranges from 380-540 feet at the base and 700 to 1000 feet at the top.
The Floodway has a maximum design capacity of 60,000 cfs. Flow into the floodway from the River is controlled by a pair of gates in the river channel 112½ feet wide each. The floodway cost $63.2 million Canadian (1967 costs) has prevented billions of dollars in damage. This year (1997) alone it is estimated that the floodway prevented close to $3.6 Billion in damages.
However, even with the floodway in operation the water level rose 24.5 feet within the City. Without the flood control works it would have risen another 10 feet to 34.5 feet.
The river peaked or crested between May 1 and May 4, 1997 with a maximum discharge of 95,000 cfs, through the City and 60,000 through the floodway.
With the river rising as it did most of the City was protected by a primary dike system, which is essentially major roads raised to a minimum elevation of 26.4 feet. Nevertheless there are some homes outside of this diking system.
In areas with low housing density, homes had to be ring-diked with sand bags or earth to be protected. In areas of higher density one continuous dike along the river bank was constructed. In total over six million sand bags were employed.
At the highest river levels precautionary evacuations were undertaken. In total about 3,000 homes (persons) were evacuated. This was done smoothly and without incident.
There were no reports of break-ins or looting throughout the incident. Evacuated City residents were out of their homes for anywhere from a couple of days to a couple of weeks.
Because 99% of the City was safely protected by the aforementioned Primary dike system, evacuees were also received from the rural areas to the south. In total the City serviced over 15,000 evacuees.
The key to a successful and timely evacuation is the development of appropriate criteria.
Our first concern was public safety and different criteria were established for different locations. Homes that were ring-diked were evacuated as soon as land access was no longer possible.
In our opinion if emergency services/First Responders couldn’t easily access the home this constituted a serious impediment to ensuring the safety of the homes inhabitants and they had to leave. In the situations with the dike on only one side and access assured on the dry-side, homes were evacuated when the river level (against the dike) rose to within one foot of the main floor of the home.
This criterion did not apply in areas protected by the large, established primary diking system.
Not surprisingly this flood involved the utilization of vast array of resources, both people and equipment. Over 3,000 City Workers, between _ to ½ of our total workforce, were engaged in flood-fighting activities.
The millions of sand bags mentioned earlier were filled using up to 4 “Sand bagger” machines, each capable of filling 50,000 bags per day.
Make no mistake 6½ million sand bags is a lot of sand bags. Laid end to end they would form a single line of bags from New York City to Phoenix. Just delivering the bags to where they were needed required 13,000 trips by truck. Putting them in place required the work of tens of thousands of volunteers.
Key to Success
1) Preparation - If you have time to prepare, use it. Prepare for the worst, you just might get it. We did!
2) Teamwork - The only way you can hope to deal successfully with a disaster like this is by working together. Not just within your organization, but amongst organizations, different levels of government, public sector - private sector, whatever. You better all be singing from the same song sheet or you’re going to have big problems.
3) Volunteers - Not enough can be said about the people who came out to help their neighbours. Without the volunteers we would have lost this war.
Mark Bennett, M.P.A., C.E.M. is the Emergency Program Coordinator for the City of Winnipeg in Manitoba Canada.
This article adapted from Vol. 10#3.
DR World Main Index | Return to DRJ's Homepage
Disaster Recovery World© 1999, and Disaster Recovery Journal©
1999, are copyrighted by Systems Support, Inc. All rights reserved. Reproduction in whole or
part is prohibited without the express written permission form Systems Support, Inc.