Sidney A. Frellson, Director, Division of Waters - Minnesota Department of Conservation
Text reproduced from Conservation Volunteer Magazine, January-February, 1963
The elements with which man has to contend are four in number; namely earth, fire, air and water. Let's consider water which, like the masquerader, is an element of many varied faces.
Water exists in the atmosphere as a vapor and is precipitated as rain and snow, dew and frost, water and ice. These forms are all chemically related, but physically they have widely different characteristics.
Most of us are familiar with the behavior of streams and lakes during the summer, but little thought is given to the winter phase of their annual cycle. With the coming of cold weather, the water in the lake or stream is gradually cooled. Temperatures within the mass of water tend toward uniformity due to continuous turbulent currents. Cold, dense surface layers sink, and the lighter bottom layers of warm water rise toward the surface.
Gradually the entire mass attains a temperature of 39.2 degrees F. (point of maximum density), and finally the surface film of water is further cooled, and when it approaches the freezing point, needlelike crystals of ice called FRAZIL ICE begin to form. On small lakes, or in the sheltered portions of large lakes, ice begins to form when air temperatures fall below freezing, but wind and wave action may prevent the complete freezing of large lakes even when the air temperature falls below 32 degrees F.
Depending upon the degree of turbulence, as well as the temperature, the crystals of frazil ice unite and form little floating masses of ice, which continue to grow by accretion of other similar ice masses. If this growth is accomplished without crushing, there results a clear thin sheet of RUBBER ICE.
The continuous ice cover now formed prevents loss of heat directly from the free water surface to the cold air so that the transfer of heat, the rate of freezing at the bottom ice surface is retarded although the ice sheet thickens steadily. Freezing often occurs very suddenly, six inches of ice cover forming in a few days. However, normal winter temperatures in Minnesota rarely produce sheets of ice more than 30 inches thick. Let's pursue this study:
- Most of us have some knowledge, varying considerably in its degree of accuracy, about SLEET, HAIL, GLAZE, FRAZIL ICE, ANCHOR ICE and the like, but as to the detailed characteristics of just "plain ice," few bother to inquire.
- Liquid water is a mixture of several molecular forms of H2O, only one of which can change into solid ice. Water, upon freezing, increases about one-eighth in bulk, but the resulting ice is less dense than the underlying water, and so it floats.
- Ice has nearly eight times as great a capacity for thermal expansion as steel. Its specific heat is only half that of the water from which it was formed, but its thermal conductivity is twice as great. While the sustaining capacity of ice is not definitely determined, tests indicate that ice two inches thick is considered safe for groups of people. Four inches thick is considered safe for light loads and eight inches thick is considered safe for loads not over 1,000 pounds per square foot.
- Railway trains have been run across ice 15 inches thick. Expansion of a sheet of ice 150 feet in width has been known to move a masonry bridge pier weighing 1,000 tons two inches out of plumb, and in other instance masonry piers on pile foundations have been pushed from two to twelve inches out of line.
- The expansion effect of lake ice, however, does not make itself felt until the ice layer is at least five inches thick. Lake shores are subjected to modification by ice action, in two general ways:
- By expansion,
- By ice jams.
- A dense snow blanket on top of the ice not only reduces the freezing rate, but prevents penetration of sunlight through the ice into the underlying water. Incidentally, this fact, according to some authorities, is one of the causes of "winterkill" of fish in relatively shallow lakes. The final result of the freezing process is a dense, water-tight sheet of floating ice ranging in temperature from a few degrees above freezing at its bottom surface to as low as 40 degrees F. at the top.
- The ice cover is normally in a state of almost complete flotation with the exception of its edges, which may be frozen solidly to the shores, projecting boulders, piers of bridges, dams, walls, or other objects to which it may have had an opportunity to attach itself. If inflow to the lake increases during the winter months, and the level rises rapidly, the center of the ice sheet is lifted.
- A sudden increase in discharge causes the water level to fall below the ice level, leaving the ice temporarily suspended. The latter condition exists for only a short time since the ability of ice to support itself in such instances is limited to spans approximating ten times its thickness, hence cracking of the ice follow, and the normal flotation level is resumed.
- Ice on a lake surface expands or contracts with the rise and fall of the air temperature, and since air temperatures have a considerable range of fluctuation in winter, the ice changes in volume. An appreciable drop of temperature causes the ice to contract, producing cracks which refill with more ice [freezing water]. When a subsequent rise in temperature produces an expansion of the whole ice mass, a tremendous force is exerted against the shore.
- If the shore is of such a nature that the ice cannot shove, it may buckle. Buckling is not uncommon, even in very thick ice sheets. Whether or not expanding ice will push up the shore or buckle depends on whether there are enough weak spots in the ice sheet to permit the release of this pressure by the buckling of the sheet or crumbling of its edges, both of these conditions being somewhat dependent on the water depth near the shore.
- If the shores are gently sloping, the expanding ice overrides them, but if the shore materials are of a yielding sort, an irregular ridge called an ICE RAMPART is likely to be formed by shoving a portion of the marginal material to a higher level, and leaving it in the form of a ridge. Such ice ramparts may be several feet in height, and may contain large boulders. Where conditions are such that ice ramparts once formed become permanent, successive shoves may build up a considerable accumulation of displaced materials, thus forming an ICE-PUSH TERRACE.
- Similarly, dropping temperatures cause contraction of the ice layer. As ice is very weak in tension, the stress is usually released by numerous tension cracks. The ice sheet already shoved up on the bank is not retracted but remains perched on the bank. Contraction of ice is usually harmless. However, the combination of alternate expansion and contraction causes a "ratchet" or "jacking" action that is more severe than either force separately.
- As contraction takes place at colder temperatures, the water that rises in the tension cracks freezes at once, filling and sealing them. When temperatures rise again, the cracks cannot close, and the entire expansion must take place at the edges of the sheet. Compression cracks due to buckling also freeze solid and are stronger than before. Each cycle of contraction and expansion shoves the edges of the sheet farther up the bank.
- The coefficient of linear expansion of ice, according to Ganot, is 0.000052 per degree of temperature rise, and thus for a 10º F variation in an ice sheet a mile long, the change in length would approximate 2.75 feet while the force exerted is probably not less than 30,000 pounds per square foot.
The writer had occasion to investigate conditions on Sand Lake, near Brittmount in St. Louis County, just north of the city of Virginia. This lake has a meandered area of 693 acres and an estimated drainage area of seven square miles. Ice action moved houses off their foundations, tipped retaining walls and left an open crevasse approximately two feet wide, two feet deep and twenty feet long in the ground. The top of the ground at this point was about five feet above the lake level. The buildings and walls affected were situated on the northwesterly shore of the lake on a sand ridge about 300 feet in width, separating the lake from a swamp area.
Some years ago a similar situation, but much more extensive, occurred along the southwest shore of Cotton Lake near Detroit Lakes in Becker County. This lake has a meandered area of 1649 acres, and an estimated watershed area of five square miles. In this instance also, rainfall and snowfall were deficient during the fall and winter period. Wide variations in temperature occurred during December and January when daily fluctuations of 25º to 30º between the high and low were frequent. The lakes in the vicinity of Cotton Lake froze over the latter part of November, opened up and froze again. Since fall precipitation was below normal and there was little or no snow on the ground, it might be termed a dry freeze-up.
There were nineteen cottages in the area examined, thirteen of which were damaged. Two had been moved back from the shore before the principal expansion force was exerted and thus were saved. Computations indicate that the total expansion of the ice, at the southwesterly end of Cotton Lake where the damage occurred, may have been as much as eighteen feet. Some buckling of the ice was found along the north shore where ridges rose to a height of 3.5 feet above the general ice level of the lake, indicating an expansion of 2.4 feet in that particular location. The foundations of some of the houses were so badly crushed that it was evident the entire ground on which they stood had been removed.
Ice, during the spring thaw, may become detached from the shores and be blown against other parts of the shore, exerting great pressures, and since these masses of broken-up ice drift with the wind, they may produce jams on and against the shore, resulting in the formation, at times, of ice ramparts.
Since these jams are wind produced, and the spring winds vary in direction, any of the less protected shores may be affected. Various other shore distortions may result, such as the change or the actual destruction of sand bars, recession of banks, formation of irregularities in shores having materials of unequal resistance on the larger lakes.
In conclusion, it appears that the effects of the "jacking" action of ice are most severe during those periods when there is little or no snow-cover, and temperatures fluctuate greatly. It is also evident that the expansive forces, which result from these wide variations in temperatures, are so powerful that it is impractical, from an economic standpoint, for the average individual to construct retaining walls or foundation of sufficient strength to resist them. The safest course is for riparian owners to be sure, where conditions are favorable for such ice action along a lake shore, that the buildings are located as high above the lake level, and as far removed from the shore as practical.
The construction of improvements on narrow ridges lying between a swamp area and the lake proper should be undertaken with extreme caution. If a retaining wall is constructed, it should have a sloping face on the lakeward side, flat enough to permit the advancing ice to over-ride it.
Trees growing near the shore are likely to be over-turned and seldom warrant costly protective works. Rip-rap is of little value unless laid on a properly placed gravel foundation and on an extremely flat slope. The conditions herein mentioned are but a few examples of the dangers which have occurred.
To all prospective purchasers of lakeshore property, let it be said that the best insurance against damage by action of the ice is to have been warned of the likelihood of its occurrence. Be sure to note the physical conditions on the property and along the lakeshore adjacent to it because the evidence of serious ice action, if it has ever been present, is there if one looks for it. Damage from this source, while regrettable, is always the responsibility of property owners.