U-Values in Autodesk “MEP Constructions for Buildings and Spaces”

[dyp4f]

https://download.autodesk.com/us/revit_mep_2016/constructions-revised.pdf

In the above document, we can see the U-values for various typical constructions (ceilings, walls, etc.) as they are considered in Revit.

However, I cannot understand how those specific U-values are calculated from those materials / layers.

Let’s go to the 1st wall, in the walls list (see page 27 of 211)

It says: Ro=0.04 m2*K/W , Ri=0.12 m2*K/W , L=0.2032 m , (Conductivity) K=0.26 W/(m*K)

We all know that the U-value calculation formula is:  U=1 / ( Ro + L/K + Ri ).

[ If we have many (n) layers, the formula is: U=1 / ( Ro + L1/K1 + L2/K2 + … + Ln/Kn + Ri ) ]

If we make the simple calculation, we will take : U-Value=1 / (0.04 + 0.2032/0.26 + 0.12)=1.06209 W/m2*K

Then, why the U-value in the Table is shown as U-Value=0.3495 W/m2*K  ?

That’s a huge difference. I can see differences of such magnitude in all U-Values in that Table.

Are they all so completely wrong ?

Has anybody else noticed those errors ?

Are internal thermal calculations in Revit being executed, based on those U-Values ? (!)

[Kevin.Lawson.PE]

Yes, there are many errors in the constructions document as well as the translation of that document into the Constructions.XML file that Revit reads (sometimes the actual construction in the Constructions.XML does not match that document).   Luckily the loads engine calculates R-Value, Thermal Mass, and transient delay factors for each construction at run-time so you can make any additions or corrections that you would like. 

The Ripple HVAC Loads toolkit comes with a constructions manager that let's you view and modify the actual Constructions.xml as you see fit. See a demonstration here: https://www.youtube.com/watch?v=qrIemhYy2jM&t=820s

If you build your constructions correctly, the internal calculators work like a charm.  The ASHRAE Journal is publishing my "Validating BIM Load Calculations" paper in July that attests to the accuracy of the load calculations.

There is a free trial for the loads toolkit, feel free to give a try for as long as you need.   

[dyp4f]

Thanks Kevin for your reply.

All those U-Values are obviously wrong.

(I will read your paper in July ASHRAE Journal)

[dyp4f]

Hi Kevin.

Congratulation for your article that was published in ASHRAE Journal, July 2022.

I wish there were a way to see a comparison between the following 3 set of data:

• The real loads measurements taken from those test constructions
• The data delivered from Revit internal calculation engine, using RTSM (from the old “Heating and Cooling Loads tool”, now depreciated)
• The data delivered from Revit external calculation via EnergyPlus engine, using HBM (from Analytical Energy Model)

[Kevin.Lawson.PE]

The RTSM (“Heating and Cooling Loads”) engine is not deprecated.  You just need to change your Revit.ini file to enable it, see instructions here: https://rippleengineeringsoftware.com/revit-heating-and-cooling-loads-in-revit-2022/

The paper addressed and compared the first two bullet points (real world and Revit RTSM), I'm waiting to address the Revit Analytical Model (HBM) method until I have a few systems type designs automated and need to compare their energy use against each other.  If you want to test out the HBM method on the ASHRAE test facility I can send you the model, just email me at Kevin.Lawson@RippleEngineeringSoftware.com

[dyp4f]

As you have showed in your article, Revit’s built-in Loads tool has 35% and 39% difference from true measured data, in heavy and light construction test facilities respectively.
Those differences are huge and absolutely unacceptable. And that’s not a Revit’s fault. It’s a RTSM’s fault. Any calculation method that miss its target by 39% is useless, as a general rule. The adoption of HBM in Revit is absolutely justified, but I want some more verification because even HBM miss the target by 8%. In my opinion, any difference more than 5% max. is unacceptable.

[Kevin.Lawson.PE]

From the paper: “It is important to note that the test facilities were designed to exaggerate the difference between the HBM and RTSM based on a study that confirmed the RSTM would always produce conservative estimates compared to the HBM and that the differences would increase as the fraction of window load increased. The large area of inefficient windows and the lack of internal loads combine to amplify the difference in the calculation procedures.”

When extrapolated to a more realistic scenario (including code minimum windows) the output of both methods tightened up to about +/- 10%. I also disagree that missing the target by 39% is useless, I’ve seen many buildings that miss the target by 300% and many buildings where loads were never calculated at all. 

The PhDs involved in the ASHRAE study couldn’t hit 5% on a simple 12’x12’ cube.  I don’t think anyone could hope to hit that on a budget with a complicated building. You should try to determine the true load on some of the buildings you’ve designed and compare them to your calculations before you form an opinion on what the maximum acceptable difference is.  Publishing your findings would be very enlightening to all of us.

[dyp4f]

Many years ago, our Professor in NTUA, during the very first lecture of the semester on Heat Transfer, said:

“You need a proven scientific methodology to precisely determine any given quantity. You must be inside a very small range around the true value. If you feel comfortable with missing the true value, by let say 40%, then there is no need for you to make all this effort to become Engineers, and for us to teach you… “

We must admit that the progress on our engineering methods during the previous 40+ years has been extremely slow comparing to the fast progress in other field, as for example in computers computing power. We have stuck in insufficient approximate methods coming from 60s and we continue to consider this as something normal.

We are just using the same ancient methods with our modern computers… with or without BIM....