RTU sequencing disaster - need advice!

Started by Bobby R. — 1 year ago — 12 views
Bobby here from Memphis, and I need help fast. Working with a retail client who has 8 rooftop units totaling about 200 tons. The building automation system was supposed to sequence startups, but something went wrong during last week's heat wave. All 8 units kicked on simultaneously around 2 PM when outside temp hit 102°F. MLGW's peak demand went from normal 280 kW to 847 kW in one 15-minute interval. This is going to cost them over $8,000 extra this month alone, plus the ratchet effect. The BAS contractor is pointing fingers at the HVAC contractor and vice versa. Anyone dealt with this kind of finger-pointing mess before?
Greg from Atlanta here. Been through this exact scenario multiple times. First thing - get both contractors on-site together, not separately. Make them demonstrate the sequencing logic step by step while you watch. Usually it's a communication issue between the BAS and the RTU control boards. Either the staging signals aren't getting through properly or there's a time delay setting that got changed during commissioning.
Jim from Minneapolis. Had similar issue with Xcel Energy customer last year. Eight RTUs, same problem. Turned out the BAS was sending stage signals correctly, but the RTU manufacturers used different communication protocols. Four units were Johnson Controls, four were Carrier. The integration was supposed to handle this but failed during high-heat conditions when response time was critical. Needed a protocol converter to fix it properly.
Pete from San Jose here. This is exactly why I always spec a demand monitoring system with load shedding capability as backup. Even if sequencing fails, the demand monitor can shut down non-critical loads before you hit catastrophic demand spikes. PG&E customers can't afford these kinds of hits with A-10 rates. The monitoring system pays for itself the first time it prevents a spike like this.
Tim from Bismarck. Check if MLGW has any demand forgiveness programs for equipment malfunctions. Some utilities will work with you if you can prove it was a system failure, not poor design or maintenance. You'll need detailed logs showing the system was working properly before the failure and documentation of what went wrong. Worth a phone call to their commercial services department.
Thanks everyone. We're getting both contractors on-site tomorrow morning. Found out the BAS programming was changed during a 'routine update' the week before the incident. The HVAC contractor claims they weren't notified of the BAS changes. Meanwhile my client is staring at an $8,000+ bill and asking why they paid for sequencing controls in the first place. This is turning into a nightmare.
Chuck from Cincinnati here. The 'routine update' is probably your smoking gun. If the BAS contractor made changes without proper testing and coordination, they should be liable for the demand charge hit. Document everything - the timing of the update, lack of notification to HVAC contractor, and the resulting failure. Most BAS contractors carry errors and omissions insurance that should cover this kind of mistake.
Derek from Atlanta. Had a similar situation with Georgia Power. BAS contractor made programming changes without testing under load conditions. All the sequencing worked fine during mild weather but failed when it was really needed. We ended up having both contractors split the demand charge cost since it was poor coordination between trades. Document who knew what and when - that's key for any insurance claims.
Juan from New Orleans weighing in. Entergy has been more flexible lately on equipment malfunction forgiveness. If you can show the system was working properly before the BAS update and the failure was due to contractor error, they might work with you. File the complaint quickly though - most utilities have time limits on these disputes. Usually 30-60 days from the bill date.
Update: Both contractors were on-site this morning. Found the issue - BAS update changed the staging delay from 3 minutes to 30 seconds between units. Under normal conditions this would be fine, but during extreme heat all 8 units were calling for cooling simultaneously. The 30-second delay wasn't enough to prevent demand stacking. BAS contractor admits the error and says their insurance should cover the demand charges. Fingers crossed.
That's great news Bobby! Make sure you get everything in writing from the BAS contractor, including timeline for insurance claim processing. Also recommend extending that staging delay to at least 2-3 minutes between units during high-demand periods. Thirty seconds is way too short for units that size. The insurance payout might take months, so document everything for your client.
Excellent detective work! This is exactly why I always insist on commissioning testing under various load conditions, not just mild weather. A 30-second delay might work in spring or fall, but summer conditions are completely different. Glad the BAS contractor is stepping up - not all of them do. This will be a good lesson learned for future projects.
Bobby, one more suggestion - consider adding a backup demand monitoring system that can override the sequencing if demand starts climbing too fast. Even if the BAS sequencing works perfectly, having that failsafe can prevent future disasters. The cost is minimal compared to an $8,000 demand spike. Sounds like you've got this situation handled well though.
Randy here from Memphis. Great work getting to the bottom of this Bobby! I see similar issues all the time with MLGW customers. The key lessons here are: 1) Always test BAS changes under peak load conditions, 2) Coordinate between all contractors before making changes, and 3) Have backup demand monitoring. This thread should be required reading for anyone dealing with RTU sequencing. Glad the contractor is making it right.