Got a welding fabrication shop here in Springfield getting killed by Ameren Missouri power factor penalties - about $1,800/month on a $8,000 electric bill. They've got mostly resistance welders and some induction equipment. The PF runs 0.72-0.78 depending on production schedule. What's the best approach for capacitor sizing with this type of fluctuating reactive load? Fixed capacitors seem like they'd overcompensate during light production periods.
Power factor correction for welding shop - capacitor sizing help
Elmer, welding shops are tricky because of the load variability. Georgia Power territory down here, we typically use stepped capacitor banks with contactors controlled by a power factor controller. Size for about 80% of peak reactive demand and let the controller switch steps based on actual PF readings. Prevents overcompensation during downtime.
Derek's right about the stepped approach. Alabama Power here - we did a 400 kVAR bank in 50 kVAR steps for a similar shop in Huntsville. The key is programming the controller with proper time delays so it doesn't hunt during short welding cycles. Also watch out for harmonics from newer inverter welders - might need detuned reactors.
Have you considered individual correction at each welder versus centralized capacitors? LG&E here in Kentucky, we had good luck with smaller capacitor units right at the welder panels. More expensive upfront but eliminates the switching complexity and reduces reactive current flow through your feeders.
Thanks for the suggestions. Leaning toward the stepped bank approach since we've got 12 welders on various schedules. Albert, can you recommend a controller brand? And what kind of time delays work best - thinking 30 seconds on, 60 seconds off to avoid nuisance switching during short welds.
Elmer, we use ABB RVC controllers here in Connecticut with CL&P accounts. Your timing sounds about right but I'd go 45 seconds on, 90 seconds off for welding applications. The longer off delay prevents oscillation when welders cycle frequently. Also set the PF target to 0.95 not 1.0 - utilities don't like leading power factor either.