NFRC Discusses Software Problems During Virtual Meeting
July 21, 2010

"Bugs" were a frequent subject of discussion as the National Fenestration Rating Council (NFRC) continued it's first-ever virtual meeting yesterday. In addition to the occasional meeting glitch, attendees learned more about some of the bugs worked out of the up-to-date versions of WINDOW and THERM software used for simulations.

During yesterday morning's meeting of the Technical Committee's Software Subcommittee, Joe Hayden of Pella, chair of the WINDOW 6/THERM 6 Approval Task Group, motioned that the subcommittee approve interim use of the software programs prior to the full completion of their validation, expected at the November meeting (CLICK HERE for upcoming meeting information).

"WINDOW 6/THERM 6 is now being used in CMAST (the Component Modeling Approach Software Tool)," Hayden said. "If we're using it there we'd also like to be able to use it in the residential program and not maintain two versions of software."

The interim approval will allow for simulation with the new software only for those products that can be currently simulated with version 5. Training for version 6 would take place online in August and September, in anticipation of a full changeover to the newer version effective as of January 1, 2011.

To begin familiarizing the committee with some of the updates to the software, Christian Kohler of Lawrence Berkley National Laboratory gave a brief presentation on some of the bugs that had been fixed in the newer version. Among those glitches was one adjustment "that results in some center-glass SHGC values increasing by as much as 0.002. Depending on a given whole product's current SHGC value, a new SHGC calculated for that same whole product with WINDOW 6/THERM 6 could increase by 0.01 when rounded to two decimal places," Kohler explained. Other bugs will have a slight impact on U-value and condensation resistance in whole products ratings.

Following the presentation, one software user expressed new concerns.

Mike Thoman of Architectural Testing spoke against the motion, noting, "I have concerns about the usability and reliability of this program." According to Thoman, "The program simply behaves erratically where things that you've set disappear on you for no good reason, things that were not issues in versions 5 of the program. Those issues make it difficult to use it reliably in a production environment."

Kohler agreed to discuss usability issues further offline to begin correcting such errors. "CMA was an addition to WINDOW 6, so that was a new feature. Until it gets tested by more people, will have bugs," he said. Otherwise, he added, "All the things that you've done in WINDOW 5 are the same in WINDOW 6."

Robin Mitchell of LBNL added, "We can't fix bugs that we don't know about so we definitely need people to tell us their problems with the program."

The group then discussed how to go about approving an interim version of the software. LBNL currently is working with version 6.3. Bipin Shah of WinBuild Inc. pointed out that, once training begins, everyone should be using the same version of the software. Others pointed out they could hardly approve a version not in current use.

"We can't approve version 6.3 when it hasn't been made available to use," Thoman said, adding, "We're not following our software approval guidelines."

Hayden agreed to add to a letter being issued to software users that "no simulator shall use WINDOW 6/THERM 6 for NFRC certification services until they have successfully completed the training."

Thoman says this addresses his concern - so nobody is using a version that's been approved without anyone using it

"Maybe what we should do is we have a review version out there for the simulators," Mitchell suggested. "We could potentially have our training and we might find more issues … we should wait until after the training, get everything resolved, so you don't have a bunch of versions floating around."

Mitchell added the group could "take the final vote at the November meeting based on the fact that the simulators have had time to test it … and everyone is comfortable with the status of the program."

However, Hayden pointed out, "Unless we force the issue we'd be in an endless loop."

The majority of attendees agreed as the motion for interim approval was passed.

The Technical Committee meeting continues today, followed by the board of directors meeting. Stay tuned to™ for more updates from the meeting.

CLICK HERE to read more about this week's NFRC meeting.

Need more info and analysis about the issues?
CLICK HERE to subscribe to USGlass magazine.