Roof snow loading has been a research area for Prof. Michael J. O'Rourke, RPI, for over 30 years. He's developed codes and standards based on observations in the Adirondacks. I expect he's making trips to Boston now to see what's happening.
As was stated by another person, the building codes identify appropriate snow loads for the geographic area. Of course, a 500-year event could exceed the code-specified required load. It's an issue of risk versus the cost for the construction. The bigger problem lies in drifting of snow due to roof configuration and adjacent structures. Some of the images shown suggest problems with connections and not the main roof members. We may have contributed to the roof problems as we increased the insulation on roofs, where most of the building heat loss occurs, allowing more snow to pile up. I don't think more insulation is bad and I also don't think a melting system is necessarily a good idea; it's a big investment for an infrequent need. Again, an issue of risk management.
As corporate architects, it's important to identify and articulate the risks associated with facilities and to identify the cost of risk mitigation. A critical facility with a flat roof would be well justified to have an ice/snow melt system as opposed to a retail facility. Of course, the person who recommended, and installed, the ice/snow melt system last year is now a company hero when roofs are failing all around him/her.
-------------------------------------------
Theodore Weidner, Ph.D., PE, AIA
Assoc. Professor
Purdue University
West Lafayette, IN
-------------------------------------------
Original Message:
Sent: 02-12-2015 20:49
From: Gary Ciccone
Subject: Snow Loading
When to shovel snow off of roofs and when to evacuate buildings.
You may think the answer to both is, shortly before the roof caves in. I tend to have roofs shoveled when we are near the design load limit and monitor deflection. Deflection in main members can cause problems with supported fixtures. I have cleared personnel only once due to covers on the lighting fixtures swinging open and a storm just starting. It was the effect of reasonable L/360 deflection of the long roof members distorting the ceiling grid. I also tend to be conservative when considering that the Utility and HVAC Engineers may not be on the same page as the Structural Engineer especially if infrastructure is likely to have been added over time. Contrarily I wouldn't want to evacuate or shovel too early and put a team at risk on a snowy roof, waist time and money doing so, and put my credibility on the line under questionable circumstances by shutting down the business for snow that would otherwise just melt off with no problems. Such is life as a Corporate Architect Facility Manger in a Biotech Plant in Andover, Massachusetts responsible for Civil Structural and Architectural Assets (CSA). I promise I won't throw a lot of corporate acrimonious acronyms at you in this article but I call this exercise CSA-CSI. This is a field probably best known as Forensic Architecture but I don't get corporate acronym credits with that. FYI this may not be PC but CYA on EH&S. I QC CD's prior to IFC not just for R&D but for all FTE's at PFE in AND.
My CSA-CSI starts by understanding the load, I like to boil the variables down to facts literally by melting snow. The volumetric weight of snow can vary greatly. . The weight of snow is based on the overall water content of the mass. Determining water content removes the variable of different types of snow. I take snow from the roof in a boiling beaker and melt it in a microwave to liquid form. In a recent check, 2200ml of snow reduced to 450ml of water equals 20% water content. A uniform coating of 36" of this snow would equal 7.2" of standing water. A cubic foot of water weighs 62.4lbs so the weight is approx. 37.4lbs per square foot. Be mindful of stratification of different layers but overall a good method to estimate weight.
Regarding the structural capabilities in a campus setting of various buildings, I focus on the weakest roofs first and determine the design load. Find the weakest link. If over the years, original contract documents have been properly archived, you can determine the actual design load for each structure or you could otherwise validate through code requirements at the time of construction assuming controlled construction. I am fortunate that most of the buildings that I manage have concrete roof decks and unoccupied mechanical penthouses but in terms of the weakest link, older buildings constructed under the 5th Edition Mass Building Code and a couple of acquired spec built warehouse buildings occupy my mind when the snow gets deep. Thankfully our building standards did not allow HVAC units on sleepers, we have dunnage racks which aid in dispersing snow loads and snow drifts. Expanding the concept of weakest link, news reports of actual building collapses in the region can also serve as an early warning system however you have to be able to discern the information. The media can sensationalize these failures along with storm severity. It can create a jittery site population looking for assurances and you can become known as the "Roof Guy" pretty quickly. Fortunately the broadcast and printed media also usually displays images of the collapsed structures which amount to prime examples of inexpensive enclosures and a barn or two build in the 1800s. Here are a few examples in New England in recent days.
Not sure this one is an actual building.
What's that, a splice in the top cord mid span?
Hundreds of dollars worth of damage!
A more reliable method I have discovered is to observe ceiling grids on upper floors. The tracks nearest and perpendicular to the wall tell the tale of deflection. The track attached to the wall is essentially supported by the floor structure below, and the parallel grids are most often supported from the roof structure above. The perpendicular track will tilt at an angle as seen in the photo below. The severity of the angle is directly related to the roof load. It beats climbing on a snowy roof anyway. If the structure is visible and accessible from underneath, hanging a plumb bob can track deflection. Remember even though deflection can be visible, shear stress is also present and very hard to detect other than by calculation.
I received an email from an EH&S colleague at another Biotech company who had recently closed a facility nearby with some photos. It was deflection causing damage to an overhead soffit detail (below)
So keep em safe and do your CSA-CSI.
-------------------------------------------
Gary Ciccone AIA
Corporate Architect Facility Manger
Pfizer, Inc.
Andover MA
-------------------------------------------