All this came to mind when we were hit with consecutive snow and ice storms this past weekend. Happily it wasn't bad as these things go. Hardened as we are to the weather in this part of the world it was no more than a minor inconvenience.
In the photos above and below you can see some of the ice and its effects. The ice isn't thick, just a few millimeters. It is enough to make a ⅛" guy wire grow to a little over ¼". The same coating went on the wire antennas and yagi yet little adhered to the towers. The weight of the ice lowered the limbs of the large spruce tree to the ground. In bad storms they're completely flattened.
This is not enough ice to cause damage except in the most flimsy of antenna installations. The trees will rebound and the antennas will continue on as normal. The only problem was that the resonant frequency of the wire antennas dropped about 6%. Ice is a dielectric that lowers the velocity factor of the wire, increasing its effective length and so lowering the resonant frequency. The yagi experienced a smaller change. Rain's impact is different in that there is not enough water to appreciably lower the velocity factor but can cause end effects which similarly lower resonance. That is probably happening here as well since ice (fresh water plus impurities) coats the insulators.
Tower manufacturers should and mostly do specify load capacity under ice conditions. When the wind blows the ice-laden tower can carry less load. Since the tower must first be able to support itself at the rated wind speed, this means a lower capacity for antennas.
First, ice lowers wind load capacity by increasing the static load (dead weight). Bending stress increases due to the weight. That is, the tower will fail with less lateral load. Second, the ice increases the wind surface area of the tower itself, lowering the wind speed at which the tower itself will fail. This is why there is less capacity for the antenna load. Third, ice increases antenna weight and wind surface area.
Consider the following tables published by Trylon (a manufacturer whose towers are commonly used by amateurs in Canada and by some in the US) for their Titan series of self-supporting towers. The first is for wind alone, and the second is for a select example of severe icing.
Trylon provides a calculator so that you can determine tower capacity due to wind and antennas alone. The calculator does not take ice into account, which is a complicating factor not easily integrated into a general formula. That Trylon excludes ice as a factor in its calculator should not be taken as license to ignore it. They would rather you consult them or hire an engineer.
Too many hams in this climate who otherwise properly engineer their antenna systems for the maximum winds they should expect to experience (wind zones) fail to take ice into account. A typical reason is they don't expect severe wind and ice to occur at the same time, thinking that is too improbable an event. I, too, have been guilty of this oversight.
Unfortunately wind and ice are not statistically independent variables. The two are often causally connected. To understand this we must detour to review some meteorology. That should illuminate the danger of ignoring ice load.
In this part of the world the typical weather pattern is for a succession of high and low pressure systems (anti-cyclones and cyclones) travelling west to east. Air moves from high to low pressure areas, and circulating vertically from the warm low to the cold high. This is obviously a simplistic description but still useful. The coriolis effect mediates the air flow (wind) such that lows circulate counter-clockwise and highs circulate clockwise in the northern hemisphere. The 3-O's mnemonic aid: a low rotates counter-clockwise in the north.
Wind speed increases as the pressure gradient increases (isobars closer together) with low and high system intensities and proximity. Air spiral outward from the high and finally spirals inward to the low. Balanced in between, parallel to the isobars and approximately on a line connecting system centres lies the geostrophic wind. The diagram at right illustrates this, though you may find it helpful to mentally rotate it 90° clockwise so that north is at the top.
Under suitable conditions the approaching low draws a warm, moisture-laden wind from the southwest, which cools as it flows north and causes precipitation. In our scenario it begins as snow accompanied by high winds, becomes rain as the cyclone centre approaches, and a decrease in wind in the eye of the cyclone. If the warm air is pushed up over the cooler air what begins as rain at altitude freezes on contact with the ground, or antennas. That is a common way to get freezing rain occurs in this region.
As the low continues eastward the winds shift to the west then the northwest as the geostrophic wind asserts itself. This wind is colder, coming as it is from a high and from the northwest. The amount of ice and speed of the wind are related since both are related to system intensity.
Therein lies the danger: the worse the icing the worse the following cold geostrophic wind. Unless the systems are moving slowly the ice doesn't melt but is fixed in place by the cold. Now you should begin to see why Trylon's ice load chart above is so important.
While it is rare for extreme winds in these weather conditions it doesn't have to be. Look very closely at those reductions in capacity. Even as I type these words 3 days following the storm the ice is still encasing the antennas, the temperature is -20° C and the wind gusts are topping 60 kph. If the ice had been thicker the risk of tower failure would be a concern. This combination of weather events can and has brought down many towers over the years, both amateur and commercial.
I had a particularly bad case of ice loading on my tower and yagi stack back in the 1980s. I was lucky that most of the ice fractured and fell off before the wind arrived. It was a tense 24 hours.
There is a reason why Maritimers tend to shorter towers and smaller antennas. It gets expensive to replace them every few years. Of course living on the Atlantic shore you can do very well indeed with less antenna. Not so here.