Working on a large commercial account in Glendale with LADWP's A-2 TOU rate. Client has consistent demand spikes during peak hours (12-6 PM) that are costing them serious money - demand charges hitting $18-20/kW during summer months. Looking at their interval data, they have a 45-minute window around 2:30 PM where usage jumps from 800 kW to 1,200 kW consistently. This appears to be related to HVAC staging, but the timing suggests there might be opportunities for load shifting. Anyone have experience with DWP's TOU structure and demand response programs?
DWP Interval Data Analysis - Peak Shaving Opportunities
Jesse - that peak demand pattern sounds like classic HVAC staging issues. APS has similar TOU windows, and I've seen 30-40% demand charge reductions just by adjusting equipment schedules. The key is understanding DWP's demand measurement methodology - are they using rolling 15-minute averages or instantaneous peaks? That 45-minute spike window might actually be hitting multiple demand intervals if they're not careful about load control.
Check DWP's demand response programs - they usually have incentives for peak shaving during summer months. PSE up here has similar programs that can offset 20-30% of demand charges if you commit to load reduction during system peaks. The interval data pattern you're describing is perfect for automated demand response. Set up controls to shed non-critical loads when you hit 1,000 kW threshold.
Jesse - I've worked with several DWP A-2 accounts. Their demand measurement is 15-minute rolling average, so that 45-minute spike is hitting three consecutive billing intervals. You're looking at demand charges on the highest 15-minute period during peak hours. Load shifting even 15-20 minutes outside the peak window could save thousands. APS has similar rate structures, and we've achieved 25-35% demand charge reductions with smart scheduling.
Gina and David - great insights. I've identified the issue - it's a central chiller system that kicks in at 2:30 PM regardless of actual cooling needs. The building automation system is programmed with a fixed schedule instead of responding to temperature or occupancy. Simple reprogramming to stage the chillers gradually between 1:45-2:15 PM should spread the load and reduce peak demand. DWP's peak window starts at noon, so shifting earlier should capture off-peak rates.
Be careful with pre-cooling strategies though. PacifiCorp territory here in Oregon - I've seen buildings that shifted cooling loads earlier but ended up with higher overall energy consumption due to thermal losses. Monitor total kWh usage along with demand to make sure you're getting net savings. Sometimes peak shaving increases total energy costs if you're not careful about building thermal dynamics.
Jesse - this is a perfect case study for TOU optimization. MLGW has similar demand patterns on commercial accounts. The key is balancing demand charges against energy costs. DWP's A-2 has different energy rates during peak vs off-peak, so shifting load earlier might increase energy costs even if it reduces demand charges. Run the full financial analysis to make sure total bill decreases. I can share a spreadsheet template that factors in both demand and energy impacts.
Randy's right about the full analysis. Xcel Energy here in Minneapolis - had a client that reduced peak demand by 200 kW but increased off-peak energy usage by 15%. Net savings were only $800/month instead of the projected $2,400. The devil is in the details with TOU optimization. Make sure your load shifting actually improves the bottom line, not just the demand profile.
Randy - would love that spreadsheet template. I'm calculating potential savings around $3,200/month based on demand charge reduction alone, but I need to factor in the energy cost implications. The client's current peak demand charges are $22,400/month during summer (1,120 kW × $20/kW), so even a 15% reduction would be significant. But you're absolutely right - need to verify the total bill impact before implementing any changes.
Jesse - also check DWP's coincident demand provisions. Evergy here in Kansas has clauses where your peak demand during system peak hours gets additional penalties. If your 2:30 PM spike coincides with DWP's system peak, you might be facing transmission charges or capacity fees beyond the basic demand charges. The interval data timing suggests this could be system peak coincident.
Charles makes a good point about coincident peaks. Chugach Electric in Anchorage has similar provisions - your facility peak during system peak hours gets charged at higher rates. Check DWP's tariff for coincident demand language. If your 1,200 kW spike happens during DWP's highest system demand periods, you could be facing transmission-level charges that don't show up clearly on the bill breakdown.
Patty and Charles - excellent point about coincident demand. I need to cross-reference the client's peak usage times with DWP's system peak data. That 2:30 PM spike is right in the middle of California's typical system peak window. If there are additional transmission or capacity charges for coincident peaks, the savings from load shifting could be even more significant than the basic demand charge reduction. This is getting more complex but potentially more valuable.
Jesse - document everything with 15-minute interval data exports before making any changes. WE Energies here in Wisconsin requires baseline documentation for any demand response claims or billing adjustments. If your optimization works, you'll want proof of the before/after comparison. Also useful if DWP questions the load shifting or if the client wants to expand the program to other facilities.
Jesse - I'll send you that analysis template this week. This thread highlights why interval data analysis is so valuable for TOU optimization. Most customers just look at their bill totals, but the real opportunities are hidden in the 15-minute data patterns. Keep us posted on your results - successful case studies help all of us make better recommendations to clients. Peak shaving is becoming more important as utilities implement more sophisticated TOU structures.