The Missing Link – Addressing Roof Ponding by Designing with an HVAC Perspective

Back to Library

Issue #11235 - February 2019 | Page #44
By J.D. Vacey

As discussed in last month’s article, “Addressing the Roof Truss Design Note: 'Provide adequate drainage to prevent water ponding.’,” by Frank Woeste and Scott Coffman, ponding is an issue not always addressed adequately.

Ponding is a design issue, a mishap, or an oversight due to a lack of experience, or possibly all three. Knowing the potential problem, the question is: do we plan for the worst and hope for the best or do we just plan? It’s sad to think that with all the intelligence and technologies in our industry, communication and integrated design are still our biggest hurdles.

Integrated design and building is a must in today’s building systems. With the fast pace of changing technologies, automation at our finger tips, and lots of experience, the resistance to change has to end. Design integration is a viable solution to cross training while we complete projects, work, and interact with each other, allowing us to share valuable information and gain even more.

By Design or Field Created – Not So Golden Ponding

Let’s consider an HVAC perspective directly. An air conditioner in an average relative humidity climate will produce approximately 3 pints of condensate per operational hour for each ton of air-conditioned space. That means, 20-ton unit x 3 pints/hour = 60 pints (7.5 gallons) per operational hour. And of course, those numbers are even higher in states such as Florida that are not in the “average relative humidity” zone. Even the condensate can present a significant issue, as it produces its own ponding around the units. Mis-leveled units or poorly installed roof drains can cause a serious pond in a matter of hours! A gallon of water is 8.34 pounds and most HVAC units operate 24/7 (7.5 x 8.34 x 24 = 1501 lbs). You can see how much deflection could occur due to a mis-hap.

Maintaining a deflection requirement to avoid any roof ponding issues is not an easy task. But neither is moving a 2000# mechanical unit – once it’s set in place, you’re too late. So let’s look at it in the design phase and not during the build phase – remember, we are structural fabricators! The expectation of us to design it in three to five days, build it in two, then wait to deliver it in three weeks to a month is unrealistic. Realistically, that three weeks’ time allows Component Designers more time to collaborate with the Architect, the E.O.R, and Mechanical Engineer. Why, you say? Because the same thing that’s missing on all three sets of plans might not have been missed.

What’s missing is the actual load of the curb and possibly other accessories. What about the rest of the system, its location, and especially the type? I know you say it’s on the plans. The same Architectural plans that refer to the Structural plans which reference the Mechanical plans that then again send us back to the Architectural plans which absolutely omitted this valuable piece of information. What about the 16-gauge duct system (dead load) that will be hanging from the component? Here’s the real kicker and I absolutely love this. When you do hear back from the Building Designer, it’s either a “structural issue” and you need get with the Engineer, or when you hear back from the Engineer, “you’ll have to get with the Architect.” But as soon as something goes wrong, the Component Designer is at fault. They misinterpreted the double detailed un-dimensioned or mis-leader-ed plans sent to them. But consider this – Component Designers have years of experience interpreting Bad plans and resolving issues before the project hits the field, and yet we are still excluded from the design process.

Mis-(sed) Understood Loads – Types and Placement

 In the next image, notice below the HVAC unit is the curb I mentioned. When the Structural set gives us a 650# for load (A), a 1615# HVAC load (B), and a 950# range hood load (C), realistically you would be following the E.O.R./Structural Engineer, right? Well, he’s human too! The question is how well do you as a Component Designer know your Mechanicals?

Mechanical plans show different types of details. They always use heavy metal duct systems and heavy electric motors. The Mechanical plans in the figure give us 390# for (A); it omits the range hood load but provides more details pertaining to size and requirements. Mechanicals also supply other loads that are not in the Structural set, such as an additional exhaust fan weighing 475# which involves another curb weight of 70#. The range hood exhaust and duct plenum will be set between two trusses weighing another 327# (D), which is an additional load not on the Structural set and the Architectural makes no mention. So, one third of the information not found on Structural or Architectural plans could be overlooked. The differential loads between the Structural and Mechanical plans equals 1000#. That’s a significant deflection load changer, isn’t it! If deflection is significant around any of the equipment, the water will start to pond next to it, causing service calls and serviceability issues. Face it, who wants to work on 20- to 30-amp electrical units or any type of electricity standing in a puddle. Just to be clear, that load did not include the HVAC unit of 1615# either.

Indeed, an HVAC curb load is always misunderstood, overlooked, and omitted from any of the plans. This is due to the Building Designer leaving it up to the builder and mechanical guys to work out. This is another reality of more misinterpreted loads omitted from our Component plans. The simple explanation of a curb is an adaptor between the roof and the HVAC Unit. Simple, right? Well, that depends on the type, and most plans will show a framed wall detail which can be loaded to our Components under Structural loads section 1603 of the Florida Building Code (pay attention to Appendix A for material weight as well).  The code states you can use a 20 psf loading for framed walls. The other detail, which you can probably already guess, is usually right next to the framed wall detail stating the unit manufacturer to supply curb. This is the one to be cautious of – they’re 14” or 24” and usually are made of 16-gauge sheet metal weighing 300# to 500#. That coupled with the overlooked 1000# and you could have the making of ponding!

Mis-(sed) Locate(-ion) – Placement of Load Issue

The next image shows two examples of insufficient information. The HVAC may be either incorrectly located or incorrectly loaded. Now we must make a call to find out, while construction is going on, and the Building Designer is already on his newest or third project since! What happens when the field misinterprets the location and places that large equipment on the wrong truss? Why don’t we get a field report? With all the technologies at our disposal, we are still not living up to our full potential in this industry. I can have a door bell send me a photo of a stranger walking up to my door while I’m visiting the Glacier National forest. But I can’t get a field crew to send me a locate dimension 20 blocks or even the next town over. You’re joking, right? The process of this industry has been broken for decades. When do we plan to fix it?

Here’s a good example of the difference of a manufactured curb, the boxed-out girder placement reflects concentrated load type. FYI, notice the extra loads that can be added to the unit but not passed on during the plan phase. Most designers that think the compressor side of the HVAC unit is the heaviest side of the unit, but they would be wrong for two reasons: the additional 110# load for the copper evaporator coil and the additional liquid weight of refrigerant itself, which can range from 20# to 30#, not only due to unit size but based on its process within the system cycle as well. Refrigerant is a vapor when it enters the compressor and the condenser coil is not as thick to perform its heat exchange stage. Which means that side of the system can be 150# less. Now look at the girders themselves. If you load the smaller one, that load will transfer to the larger ones, and we all know that a concentrated load is more stressful than a pound per linear foot load (PLF). This will affect the deflection!

Moving Loads – Field-Proofing Load Displacement

From a Component Designer standpoint, do you anticipate field problems due to poor plans, or do you follow the plans without question? How well do you know your customer? Are they flexible or are they always in a hurry? The smart ones will listen to their manufacturer’s experience and allow us the leeway to make adjustments. The really smart ones give us the plans before the Structurals get done – this way we can save them thousands, and let’s face it we make the Engineers’ lives easier. Personally, field proofing has become a way of design in my Components, being a former framer back when trusses were the new technology and we all gasped at the notion of it. With all those previously discussed loads, there is a failsafe when we design and analyze our Components; it can be time consuming, but the outcome speaks for itself. Analyzing a moving load to bump up your plating size and lumber grade can be beneficial if the possibility of a field error or load placement issue can occur. The cost difference is minimal, and you should keep an eye on the design as you go, not to overdo it. Look at it as a type of insurance for your customer as well as the field crew’s safety – some of them are new too. Remember when you started? The $10 to $20 difference can be made up somewhere else. Realistically, Component Designers would not have to take such drastic measures if we were credited for our accomplishments. After all, we are professionals too. Integrate your design with us, we can help! Or you could consider the cost of what the repair would be if the HVAC unit was placed on the truss that wasn’t loaded for it. Maybe it’s a key aspect to a well-managed deflection for the entire project.

Define a Well-Managed Deflection

It’s careless to think Component Designers are not multifaceted professionals who go above and beyond their duty to supply their clients with the best possible design and fabrication solution to their project. While addressing some of the long-term effects of this mis-managed industry, some of us have developed safe guards to protect everyone involved. For example, the next image shows RTU 1 spanning four trusses and, with E.O.R. approval, it can be loaded with PLF load. RTU 2 only bears on two trusses and, if the framer does his job correctly, will bridge the third truss on the right to help support the HVAC unit. Let’s do a “JD” Reality Check and realize that the field could misplace the units onto the next truss over. What effect would that have and what type of repair would it take to fix such an error? Therefore, field-proofing jobs to avoid this frequent scenario is always beneficial. The minimal cost of a few bumped-up plate sizes and some lumber upgrades far outweighs the cost and time of any repair. You don’t do that for your Client? Why not? I can’t tell you how many times a project manager or super has called me with this type of issue. I kindly explain how I anticipated such an error and direct them to the Engineering paragraph that shows the loads that cover them. Have you ever considered how much time and money that type of forward thinking saves your customer? Now ask yourself, whether you’re a Builder, their Designer, or the Manufacturer, wouldn’t it be better if we collaborated more between the office and field? Then this would not need to take place at all! I’ve seen many field conditions that did not match the plans, and I’ve always wondered if that change ever made it to the Component Designer before fabrication, especially when the component package has been sitting on the ground in the weather blocking deliveries and sub crews for three weeks to a month.

32” O.C. Spacing with a Hurricane Wind Load to Parapet Wall

The final image shows a total project deflection of 3/16”,which ranges from 1/2” to 11/16” and only 1/8” difference between individual trusses, with loads sporadically placed and some failsafes in place. Let’s finish our design criteria by adding the 65# PSF wind to our design per the E.O.R. instructions. We must create two new load catagories by copying the original dead load already supplied in our MiTek software. We create a load case 25 for internal direction and load case 26 for external direction. Now we have to take the PSF Load the E.O.R. has given us and turn it into a PLF Load for application. The formula is: PSF Load (65#) times the spacing (32”) equals 174 PLF.  The reason this type of loading does not affect the deflection limits is due to the lumber and plate increases it demands for horizontal shears that are created.

I hope you can see why we need to collaborate more and truly integrate our industry at the design phase and not just in the field phase. The forward thinking of integrating these multifaceted cultures can bring more value to any one office. It’s not a new concept just one we refuse to use. The missing link is not just the fact that the building industry looks at Component Manufacturers as a vendor only! It’s the sharing of knowledge based on education level learned from a book to experience levels that cannot be learned from a book, only from years of hands-on work. Linking the field with office to office communication can advance any individual’s or team’s experience no matter what level of education they have.

J.D. Vacey

Author: J.D. Vacey

Truss, Wall, Component, and Mechanical Designer

You're reading an article from the February 2019 issue.

Search By Keyword

Book icon Issuu Bookshelf