Water-Supply Paper 2499

U.S. Department of the Interior
U.S. Geological Survey

Summary of Floods of 1992

March 11, 1992, Ice-Jam Flood in Montpelier, Vermont

By Jon C. Denner and Robert O. Brown

The ice-jam flood on March 11, 1992, caused the largest inundation of Vermont's State capital since the flood of 1927. Unlike the 1927 flood, which occurred because of intense rainfall and excessive runoff, the high water levels of March 1992 were generated by ice-induced backwater of the Winooski River. Although the areal extent of an ice-jam flood is small, the consequences to a community may be severe. That was the case in Montpelier when the Winooski River overflowed its banks and inundated the downtown area.

Pre-Breakup Conditions in the Winooski River Basin

In mid-February, the Winooski River and North Branch Winooski River in the Montpelier area were covered with solid ice and snow. Streamflow on unregulated streams was low because of consistently cold temperatures. On February 18, a discharge of 232 cubic feet per second was measured in the reach upstream from the streamflow-gaging station on the Winooski River at Montpelier (fig. 30). The discharge measurement was made through ice cover using a current meter. Backwater attributed to ice effect, as determined by the discharge measurement, was 1.21 feet. The average ice thickness in the cross section was about 1.7 feet. After a brief thaw and light rains on February 19, temperatures remained seasonably cold through the first week of March.

A moderating trend began as a storm system developed over the mid-Atlantic Coast and moved in a northeasterly direction along a cold front bringing rain and above-freezing temperatures to Vermont. When ice-covered rivers, such as the Winooski River, are subject to mild weather, the following two processes are likely to occur: (1) increased runoff by snowmelt and rains can result in increased uplift and frictional forces applied to the ice cover, and (2) increased heat input to the ice can reduce its strength (Beltaos and others, 1990, p. 39).

Although snow depth had diminished to 2-4 inches in the Montpelier area, a considerable snowpack remained in the higher altitudes on March 10. Snow-course data, collected on March 3 at an elevation of 1,300 feet in the headwaters of the Winooski River Basin, indicated an average snow depth of 18 inches and a water depth equivalent of 5 inches (unpublished data on file with the Bow, New Hampshire, office of the U.S. Geological Survey).

Late in the evening on March 10, a low-intensity rainfall commenced over the Winooski River Basin and continued overnight. By 0600 on March 11, the storm dropped about 0.60 inch of rain. An additional 0.20 inch accumulated throughout the day on March 11. About 0650, local police reported an ice jam near the Washington County Railroad Bridge, west of Pioneer Street Bridge (fig. 30). The jam subsequently released, and the ice and water surged downstream. At 0700 another police report described an ice jam near the Bailey Avenue Bridge, and flooding was reported on State Street (Times Argus, March 15, 1992). Within less than 1 hour, downtown Montpelier was inundated to a depth of 2 to 5 feet. A formal state of emergency was declared by the Governor of Vermont at 0900.

Ice from the 1.5-mile section of the Winooski River downstream from the confluence with Stevens Branch probably caused the ice jam in Montpelier (Federal Emergency Management Agency, 1992). A surge of water released from the upstream ice jam may have triggered ice breakup along the Winooski River in Montpelier. The ice run stalled at a bend in the channel about 300 feet downstream from the Bailey Avenue Bridge. Stage data recorded at the streamflow-gaging station downstream from the major ice jam indicated substantial backwater in this reach prior to the formation of the jam (fig. 31). Another ice jam downstream from the streamflow-gaging station, possibly in Middlesex, may have caused the backwater. Channels affected by backwater typically have a marked reduction in water-surface slope and a decrease in flow velocities. These factors may have contributed to the stalling of the ice run and its subsequent ice-jam formation on the Winooski River.

After the initial ice jam formed near the Bailey Avenue Bridge, additional ice fragments continued to arrive from upstream. The jam thickened primarily because of underturning of ice blocks. Rising water levels caused large ice fragments to become entangled with the low steel of the Bailey Avenue and Taylor Street Bridges. Ice was stacked against the upstream side of the Washington County Railroad Bridge, but about 1.5 feet separated the Main Street Bridge from the ice pack. Bridges and other structures can contribute support to ice jams. The ice jam extended about 1 mile along the Winooski River, its toe (downstream end) was located near the Bailey Avenue Bridge, and its head (upstream end) was upstream from the Granite Street Bridge (Federal Emergency Management Agency, 1992).

Montpelier Flooding

Flooding of the downtown area was rapid. The ice jam formed about 0700, and by 0800, Main, Elm, and State Streets were inundated. Most office workers, merchants, and residents had little warning of the impending flood. Some waded through thigh-deep water in parking lots only to find their vehicles stranded. Hundreds of people were evacuated by local and State police, fire departments, and private citizens using an array of small watercraft. Fortunately, the flood occurred during daylight; otherwise, the rescue operations would have been more difficult.

Flooding was observed first on State Street in a low-lying area near the confluence of the North Branch and Winooski River. The high water levels on the Winooski River, resulting from the ice jam, created backwater on the North Branch Winooski River. The North Branch overflowed sending floodwater onto State Street. Ice cover along the North Branch was uplifted but remained intact as the water level increased. Downstream movement of ice on the North Branch was prevented because the ice pack on the Winooski River blocked its outlet. Furthermore, ice cover on the North Branch lodged against the Langdon, Rialto, and Washington County Railroad Bridges.

Flow on the North Branch Winooski River is regulated by the Wrightsville Detention Reservoir (site 3, fig. 32), located 4.2 miles upstream from the confluence with the Winooski River. The earthfill reservoir, constructed for flood-control storage, was completed in 1935. The effectiveness of the structure was documented during the 1936 flood; the reservoir contributed to reducing the flood crest and potential for flood damage in Montpelier (Denner, 1991, p. 539). As originally designed, the reservoir was an uncontrolled, self-regulating detention basin. Outflow was dependent on the capacity of the outlet opening near the base of the dam. Since 1985, a hydroelectric-generating station has operated at the reservoir outlet. When the reservoir stage is below 635 feet, discharge is through a conduit leading to the generating units. Water levels higher than 635 feet flow out an uncontrolled conduit; discharge at high stages, then, is a direct function of the reservoir stage.

The streamflow-gaging station on the North Branch Winooski River (site 2, fig. 32) is 0.8 mile downstream from Wrightsville Detention Reservoir (site 3, fig. 32). Recorded data showed discharge on March 11, from midnight to 0500, at about the minimum-flow rate of 30 cubic feet per second (fig. 33). At 0600, discharge increased to 215 cubic feet per second as a result of powerplant operation. Powerplant operation continued until shutdown at about 1015. Meanwhile, the water level at Wrightsville Detention Reservoir was increasing (fig. 34). Outflow to the uncontrolled conduit began at about 1100 when the reservoir stage exceeded 635 feet. Discharge on the North Branch increased during the afternoon; a discharge of 563 cubic feet per second was recorded at 1715, approximately when the ice jam released on the Winooski River. Discharge probably was slightly higher at the mouth because of additional inflow from streams between the streamflow-gaging station and Montpelier. A maximum discharge of 842 cubic feet per second occurred at the North Branch Winooski River streamflow-gaging station on March 12, at 0530.

Discharge on the North Branch Winooski River alone was too small to account for the rapid flooding of Montpelier. Estimated streamflow on the Winnooski River was about 3,000 cubic feet per second during the period of inundation; thus, the Winooski River was most likely the major source of floodwater in the downtown area.

High Water Levels Resulting from Ice Jam

A major consequence to communities during ice-jam flooding is the high water levels attained behind the ice dam. An important constraint to the size of an ice jam and thus the maximum water level is flow diversion around the ice jam (Beltaos and others, 1990, p. 77). After the Winooski River and North Branch Winooski River overflowed onto the flood plain, water was free to move around the ice jam. The ice dam at the Bailey Avenue Bridge was bypassed on the north bank. Lower State Street, in effect, became a spillway for the ice dam. Overbank flow on the flood plain reconverged with the main channel about 200 feet downstream from the Bailey Avenue Bridge (fig. 30).

The water-surface elevation in the impoundment was relatively stable during most of the flood. Water levels, based on onsite inspections, ranged from 523.4 feet at 1055 to 524.3 feet at 1630. A maximum water level of 525.1 feet (0.9 foot below the current Federal Emergency Management Agency 100-year flood elevation) was determined from high-water marks found at the Federal building on State Street. The maximum elevation probably occurred during the surge of ice and water at about 1700. In comparison, the 1927 flood crest of 533.9 feet exceeded the 100-year flood elevation by 7.9 feet.

Ice-Jam Release

Between 1430 and 1500, a section of the toe of the ice jam dislodged as a result of high flows and intervention by construction equipment; a crane operating on the left bank dropped a steel beam on the ice fragments while excavators pushed blocks downstream. The ice jam redeveloped, however, when upstream ice fragments moved downstream. Ultimately, the ice jam was pushed out by a major surge of ice and water. The surge originated in the steep section of the channel between the Stevens Branch confluence and the upstream dam at Levesque Station. An ice jam at the confluence broke up at about 1615. The flood surge traveled downstream to the Bailey Avenue Bridge, causing breakup of the ice jam there at about 1710.

As the jam moved out, ice damaged the right truss of the Washington County Railroad Bridge, thereby causing the bridge to fail. The bridge was driven off its center pier by a large mass of ice and snow that had accumulated over the winter as snow was removed from the city streets and dumped into the Winooski River. The mass remained lodged against the bridge after the water receded (Federal Emergency Management Agency, 1992).

When ice jams release, water in storage discharges, and sudden increases in water levels and velocities are generated downstream. The maximum gage height and discharge recorded at the Winooski River streamflow-gaging station downstream from the Bailey Avenue Bridge were 15.71 feet and 11,500 cubic feet per second, respectively, at 1730. The maximum discharge had a recurrence interval of about 10 years (10-percent chance in a given year). The 1992 flood maximum was much smaller than the 1927 flood maximum. A maximum gage height of 27.1 feet and a maximum discharge of 57,000 cubic feet per second occurred during the 1927 flood; the recurrence interval was greater than 100 years (1-percent chance in a given year).

The duration of flooding in the downtown area was about 11 hours. After the ice jam released, floodwater quickly receded from the streets of Montpelier. The surge of ice and water traveled downstream causing overbank flooding in the fields between Montpelier and Middlesex. The arrival of sharply colder weather later in the day on March 11 reduced runoff and thus lessened the potential for more flooding in the Montpelier area.

Despite backwater from ice-affected streamflows at the streamflow-gaging station on the Winooski River, recorded stage data provided valuable information on the ice-jam flood. The gage-height plot (fig. 31) illustrates streamflow trends. Stage and discharge increased during the early morning on March 11. The sharp spike at 0715 shows the surge following the release of the ice jam upstream from Montpelier. Discharge decreased after the ice jam formed downstream from the Bailey Avenue Bridge. Between 0730 and 0800, the flow by the streamflow-gaging station was relatively stable, probably because of water being retained by the ice dam. The upward trend after the inundation of the downtown resulted from flow bypassing the jam and increased runoff in the basin. About 1500, heavy equipment dislodged some ice in mid-channel. The rapid drop during this period probably was not caused by a reduction in discharge but instead may indicate a reduction in backwater. By 1600, the ice jam redeveloped, and flows continued to increase. The flood wave that ultimately caused the ice jam to fail arrived at about 1700. High flows, related to the dynamic breakup, are represented by the sharp upward trend. The recession after the maximum at 1730 shows decreasing discharges under mostly open-water conditions.

Flood Damage

The downtown commercial district of Montpelier received severe damage from the flooding. Water levels were 2 to 3 feet above the main-level floors in many businesses. Flood damage consisted primarily of destroyed inventory, machinery and equipment, and records and utilities housed in basements and on main-level floors. Buildings, streets, sidewalks, and a railroad bridge were damaged (Federal Emergency Management Agency, 1992).

Martial law was declared in Montpelier on March 11, and only business owners and displaced residents were allowed in the city. Cleanup efforts were hampered by extremely cold weather and light snows. The first priority for many property owners was to pump out basements and to repair heating and utility units because subfreezing temperatures could have further damaged properties. More than 200 automobiles were damaged or totally destroyed by floodwater. Some vehicles, not towed to heated garages, sustained more damage because engine blocks and transmission cases were cracked by expanding ice.

Petroleum spills caused pollution and safety hazards. An estimated 8,000 gallons of fuel oil were discharged into the floodwater (Federal Emergency Management Agency, 1992). In addition, gasoline leaked from automobile-service stations and vehicles. These contaminants either evaporated or were flushed downstream with the high flows after the ice jam was released. However, some petroleum residue remained in buildings and soils and created potentially hazardous conditions for emergency crews. In Montpelier, the ice-jam flood caused an estimated $4 million in damage. Other flood damage totaling $1.1 million occurred in Caledonia, Orange, Washington, Windsor, and Chittenden Counties (Federal Emergency Management Agency, 1992). The President of the United States declared the flood-affected counties a disaster area (Times Argus, March 15, 1992). No deaths or serious injuries were reported.

References

Beltaos, Spyridon, Gerard, R.E., Petryk, S., and Prowse, T.D., 1990, Working group on river ice jams, field studies and research needs: Saskatoon, Canada, Nation Hydrology Research Institute, 121 p.

Denner, J.C., 1991, Vermont floods and droughts, in U.S. Geological Survey, National Water Summary 1988-89: U.S. Geological Survey Water-Supply Paper 2375, p. 535-542.

Federal Emergency Management Agency, 1992, Interagency hazard mitigation team report for Vermont: FEMA-936DR-UT.