[BLDG-SIM] Bug in Window 5.2

[Joe Huang]

Bug in Window 5.2Mike, others,



You can regard this e-mail as the "LBNL side of the story" that you've requested. The reasons for our delay in responding are threefold: (1) we needed to investigate the situation thoroughly, (2) provide a solution for any defects, and (3) several of us were tied up at ASHRAE the earlier part of this week. We at LBNL stand behind the WIndow program and look carefully into any reported bugs and mistakes.  Any imputation that we are dallying when we learn of problems, especially for nefarious political reasons (!), is unwarranted and only lowers the level of discourse. 

 

When this problem was first mentioned a few months back, the Window 5 development team was under the mistaken impression that it did not affect the calculations, since DOE-2 strips off the air film coefficients anyway. When it was explained that DOE-2 does this only for the exterior air film, and that the mistake in the interior air coefficient does affect the numbers, the development team started to look into this in more detail and to provide a fix. 

 

We determined that this is purely a reporting defect and not a calculational error in Window 5. Without going into the programming details, suffice it to say that as the results are passed to the DOE-2 W5 file, the radiative portion of the interior film coefficient got left off inadvertently, so that only the convective portion was being reported on Line 53 (0 degree C, 6.71 m/sec wind speed).

 

We also found a small reporting bug in the file produced for EnergyPlus with lines in the Gas Data section that are more than the 132-character limit in the EnergyPlus routine that reads the file. We are now making bug fixes for both of these reporting mistakes in Window 5, and will release a corrected version at the end of July. In the meanwhile, we suggest the following work-around to correct the DOE-2 W5 file produced by the current version of Window 5:

 

The interior air film coefficient is related to the center-of-glass U-factor (Ucog) by the equation  



            Hin,tot = 6.9681*Ucog .0491     (Hin,tot and Ucog both in w/m2K)

 

 

Use this equation to calculate "hin" based on the U-factor listed immediately to the right and substitute it in place of the existing "hin", which contains only the convective portion of the interior air film coefficient.



This correlation is valid for 1 < Ucog < 5.2 W/m2K for windows that don't have a low-e surface facing the room. It is quite uncommon to have a low-e coating facing the room, but in that case the radiative component of the film coefficient will be very different. Otherwise, the error for calculating "hin" based on this equation is < 0.5% from the actual value.

As an example, shown below are the original and corrected lines for a sample DOE-2 W5 file :



Original:

Overall and Center of Glass Ig U-values (W/m2-K)
Outdoor Temperature                 -17.8 C      15.6 C      26.7 C      37.8 C
Solar      WdSpd  hcout hrout  hin
(W/m2)     (m/s)     (W/m2-K) 
  0        0.00   4.00  3.39  2.68  2.14 2.14  2.26 2.26  2.33 2.33  2.51 2.51 
  0        6.71  30.84  3.23  2.84  2.74 2.74  2.83 2.83  2.90 2.90  3.14 3.14 
783        0.00   4.00  3.71  1.65  2.14 2.14  2.26 2.26  2.33 2.33  2.51 2.51 
783        6.71  30.84  3.32  2.29  2.74 2.74  2.83 2.83  2.90 2.90  3.14 3.14 



Corrected:

  0        0.00   4.00  3.39  2.68  2.14 2.14  2.26 2.26  2.33 2.33  2.51 2.51 
  0        6.71  30.84  3.23  7.32  2.74 2.74  2.83 2.83  2.90 2.90  3.14 3.14 
783        0.00   4.00  3.71  1.65  2.14 2.14  2.26 2.26  2.33 2.33  2.51 2.51 
783        6.71  30.84  3.32  2.29  2.74 2.74  2.83 2.83  2.90 2.90  3.14 3.14 

 

Correction:

        hin =   6.99681 * 2.74 0.0491     =  7.32



Joe Huang

Simulation Research Group

and on behalf of the Window and Daylighting Group

Building Technologies Department

LBNL





In case there are any awk users out there, I've attached a simple awk script (w4cor.awk) to make these corrections to exsiting DOE-2 W5 file with these reporting errors.  This awk script will automatically change the label on the first line to Window 5.2r (to prevent future confusion), and then make the correction just described. The syntax for running awk in a DOS window is 

   

       awk -f w4cor.awk [input file name] > [output file name]

 

Publicly available versions of awk (or gawk) can be found on the Web by searching for "PC awk downloads". 



w4cor.awk script :



{if (filecount > 0) filecount += 1 } 
{if (substr($0,1,10) == "Window 5.2")
   {
      substr($0,11,1) = "r" 
      filecount = 1
   }
}
{if (filecount == 53)
    {
       begline = substr($0,1,30)
       endline = substr($0,36)
       logUval = log($6)
       hincor = logUval*0.0491
       hin = 6.9681 * exp(hincor)
       printf("%s%5.2f%s\n",begline,hin,endline)
    }
 else
    print $0
}
----- Original Message ----- 

  From: Mike Tillou 
  To: BLDG-SIM at gard.com 
  Sent: Friday, June 17, 2005 6:02 AM
  Subject: [BLDG-SIM] Bug in Window 5.2


  Hi All, 

  Anyone using Window 5 (any version) to create custom glazing for DOE2 simulations should be very careful.  There is a bug in the output feature of Window 5 that incorrectly reports the window film coefficients used to calculate the center of glass U-value.  If you use this data in a DOE 2 simulation the calculated center of glass U-value may be off by as much as 40%.  BE CAREFUL and check the output reports to verify glazing performance if you are importing Window 5 performance data.  Apparently Window 4 does not have this problem, it worked fine when I tried it.  

  I have been told that LBNL is aware of the problem but to date has not resolved the issue.  

  I have only heard one side of the story (and would be happy to hear the other) but it seems like this is another chapter in the ongoing saga of DOE2 vs Energy Plus.  I am not saying this is the case, but it is disturbing to think that software bugs in publicly funded software are not getting fixed for political reasons.

  Unfortunately those of us trying to use these tools and can't because of issues like this suffer most. 

  Sincerely, 

  Disgruntled Modeler. 



  Michael Tillou, PE
  etc Group, Inc.
  PO Box 7, Williamstown, MA 01267
  ph. (413) 458-9870  fx. (413) 458-9875

    


==================
You received this e-mail because you are subscribed 
to the BLDG-SIM at GARD.COM mailing list.  To unsubscribe 
from this mailing list send a blank message to 
BLDG-SIM-UNSUBSCRIBE at GARD.COM


------------------
ATTACHMENT REMOVED
------------------