Vehicle Mishap:  Waterbury Reservoir, VT

 Red=accident site, Green=parking area, Yellow = big rock, Blue=sample location near the dam. Imagery-© 2012 TerraMetrics, Map data ©2012 Google.

Where the truck broke through 2-1/2 days earlier. The darkest area is the west end of the hole. The more grey ice with black ice chunks is the front of the hole. The water is a little over 50 feet deep.

Date and Time: January 17, 2012 in the evening after dark. 

 Summary: A new F150, extended cab, truck with a plow and two people broke through while plowing fishing shanty access roads on the ice on the East Arm of Waterbury Reservoir, VT. Both people successfully exited through the driver's side window with the last person leaving as water started to pour in the window. They estimated the time for the the truck going under at about 30 seconds.

The Ice:  I sawed out blocks of the ice sheet from several places on the East Arm and one place near the dam (see map).  The underlying black ice was large grain ice (type S1) with a thickness ranging from 4.7 to 5.5".  The ice on top of the black ice was snow ice (frozen snow slush).  In some places the slush had not frozen all the way, leaving a layer of unfrozen slush between the snow ice and the underlying black ice (layered ice).  This was found in about 20% of the places I drilled.   There was dry snow on top of the snow ice in some places which served as a  good insulator for any underlying slush layer. 

It looks like the black ice got submerged by snow that fell on January 12 and 13 (0.47" water equivalent in nearby Montpelier).  This established a slush layer which froze to an 1-3" thick in places where it was not insulated by dry snow.  The worst ice in the area around the breakthrough  was comprised of some snow over an inch of snow ice over an inch of slush over 5" of black ice. 

The following picture shows an example with a thicker snow ice top layer than was found close to the breakthrough.

An example of the ice structure: about 5" of black ice, 1" slush and 3" of snow ice (frozen slush). The camera parallax makes the measurement look smaller.

The breakthrough thickness for a 5,000 lb truck+plow is about 5" of good quality black ice (no cracks) at 32 degrees (Note 1). There was not not much excess strength in the ice sheet, especially if the black ice layer had to carry all the load.  The Minnesota DNR recommended minimum thickness for this heavy a truck is 12" to account for  imperfections in the ice, other loading factors and a reasonable safety factor. The Mn DNR and CRREL suggest counting good quality snow ice at half its thickness. A five inch black ice base would need 14" of good quality snow ice on top to meet their recommendation (19" total).

 Weather History:   Some of the fishermen told me that the ice developed in stages starting at the east end of the Arm in late December.  Based on the temperature record and the uniform black ice thickness it looks like the rest of the lake froze over in the first week of January.  This is late for this lake.  For reference, in the Champlain Valley the ice season started about 2 weeks later than average.

The day of the breakthrough the temperatures rose through the '30's during the day with light freezing rain and plain rain.  By the evening the snow on the ice was wet and was expected to harden up in the cold temperatures expected the next few days.  That may have factored into the decision to plow then rather than waiting until a later date.

 Some of the Circumstances Leading to the Mishap: The driver drilled several  places on the area he planned to plow.  He found 9" to a foot of ice. He was plowing to make the access to the shanties easier, as has been done in past years.  They plowed a road north east from the launch area and then to a large rock outcrop on the north shore to the west without incident (the snow covered rock is visible in the  picture after the google map).  He was plowing back to the ramp at about 10 mph when there was a loud bang, followed by another,  which stopped the truck dead and bounced  both people onto the steering wheel or dash. 

 Breakthrough and Exit Details: I talked to both the driver and passenger. They both said it happened fast and they were lucky to be alive.  Some of the details are not clear but generally what happened is  they quickly realized the truck was sinking.  One of them broke the tempered glass drivers side window (with his elbow) and went out the window on to the intact ice beside the truck.  He then pulled the other person out as water started to pour through the window.  They estimated the truck went under in about 30 seconds.  Luckily this was much longer than the 7 second floating phase reported in a test with a 1 ton truck with a plow (note 2).   The water depth was over 50 feet.  Dr Giesbrecht reports that, once the vehicle goes under the surface, the odds of surviving are poor.

 Aftermath and Recovery:  The driver notified authorities and his insurance company the next day.  His coverage will pay for the vehicle and recovery. It looks like vehicles with comprehensive coverage will cover your vehicle and its recovery if there is not a exclusion that applies. If you have an older vehicle with just  liability coverage, you will most likely have to pay the recovery cost for the vehicle and be left with the salvage value of the vehicle.   If you drive on ice, it might be worth talking to your insurance agent to see what your policy covers.

The recovery plan was to remove the vehicle soon after the ice went out.  I was told this allows a more straight forward recovery.  The fuel and oil in the vehicle do not tend to escape until the vehicle is moved. It is easier to manage any leakage on water than trying to manage it under ice.  The recovery was completed on April 4th by Champlain Divers. The team of four, using a workboat attached flotation bags to lift the truck off the bottom.

Work/Dive Boat. Lots of air bags, lots of air,chain and a plan for how to do the recovery.. Sonar image of the truck on the bottom. The GPS location was established on the ice just after the accident and it was used to get into the immediate area. Adding and adjusting the airbags near shore to roll the truck onto its wheels so it can be pulled out.

It was then floated into shallower, more protected water for the attachment of more bags to put it in a wheels down position.  The truck was then pulled onto shallower water where the cable from a tow truck was attached to pull it the rest of the way out. The owner's wallet, fishing gear and truck cap were all where they were when he last saw them.

Waterbury Truck looking pretty good as it come out of the waterIn 2009 this truck was pulled to shore through a 600 foot slot cut in the ice over water about four feet deep. The team was resourceful but learning as they went. Paying professionals is probably worth the cost ($3,000-5,000??) in terms of salvage value of the vehicle, especially if it is new (as both these vehicles were).

A happy truck owner and the recovery team

 Thoughts and lessons:

The Ice:  Moderate or heavy snow on top of relatively thin ice  can be treacherous.  This condition played a big part here as it did at Lake Dunmore in January 2010. At both accident scenes, there were areas that were solid and thick mixed with others that were thinner and with a slush layer leaving only the black ice layer providing meaningful support.  This condition is more likely in the first 4-6 weeks after the ice comes in but, in theory, it can occur anytime. It is likely to develop whenever a few inches or more of snow falls on relatively thin ice. This ice condition fairly tricky to access.  Its variably means you have to look a lot of places to see what is going on and identify potential trouble areas.  The best way to assess the ice is with a drill and ice saw by cutting a small block loose and pulling it out for inspection.  Since few people have a saw, a sharp and carefully used hand auger, will give a pretty good idea of the presence of partially saturated snow ice (cuts faster and feels a little different than hard ice) and slush layers (the drill binds a little and drops when it gets to the slush layer).  A drill does not easily discriminate between hard, saturated snow ice and black ice although you may be able to see the difference by looking into the hole.  A power auger is likely to be considerably less sensitive than a hand auger.

I do not know how far from shore the test holes were made before the accident.  My testing  found the thickest ice nearest to the parking lot.   In this case, more test holes over the entire area to be plowed and closer examination of the nature and variability of the snow ice/slush above the black ice would have most likely led to waiting for thicker ice before putting a truck on it.

The decision to plow:  With the benefit of hindsight, it is very easy to question the decision to plow when they did. The problem is we often don't have hindsight when we are making these decisions.  The better we understand what we are dealing with, the better the decisions we will make.   Cars and trucks present an especially high mortality risk on ice. This makes decisions about ice travel in/on vehicles especially important. Related to this, it is human nature to have a lack of appreciation for just how bad things can be if they go wrong.   One of the reasons for this website is to give some understanding of the ground truth as a way to share the insight gained from past accidents.

Exiting: The driver and passenger made the right decisions on getting out of the truck:

  • They already had their seatbelts off
  • They managed panic well enough to be successful in saving themselves.
  • They chose to go out the window rather than opening the door.  If they had tried and been successful, opening the door would have filled the cab before they could have gotten out.  The chances of surviving after the truck went under were slim, especially at night in 50 feet of water.
  • They successfully broke the window in spite of not having a proper tool for the job.  Tempered glass is very difficult to break and most people would be hard pressed to do it with their elbow or feet.
  • They got out fast with the first man out helping the second. 

There are lots of stories of trucks going through when plowing snow on lake ice.  If you are considering doing this you you might consider reading the report linked to Note 2 below.  They found that putting two 55 gallon drums on top of the plow increased the floating phase from 7 seconds to over 61 seconds in their test (see the picture and notes on the last couple pages).  Also, anyone who plans to drive anything on the ice should watch Rick Mercer and Dr Popsicle or read this article.  If you ever consider putting a heavier plow truck on the ice, watch this first. It is hard to believe how fast a heavy truck skinks (2-3 seconds!)

Speculation on the breakthrough: The driver thought the initial 'bangs' might have been when they hit a crack or pressure ridge.  I did not find any evidence of a ridge or wet crack although the ice was snow covered.  An alternative possiblity is the plow might have broken through the snow ice crust creating the first bang.  The second bang might have been the breaking of the underlying black ice.



Note 1) The break through thickness estimate is based on Fransson-Ice Engineering - The formula can be found on Page 19.  I assumed 0.8 MPa flexual strength for the ice, an all up vehicle weight of 6000 lbs and a 1.5 meter load radius

Note 2) Exit Strategies and Safety Considerations, for Heavy Machinery Occupants Following Ice Failure and Submersion, G Giesbrecht and G McDonald, 2010. See page 7.