Working on an audit where about 30% of the bills over a 24-month period are estimated readings. The utility (MLGW) says they had meter access issues and equipment problems. How do you handle estimated bills in your analysis? Do you exclude them entirely, try to interpolate actual usage, or take some other approach? The estimated vs actual swings are throwing off my demand analysis.
Dealing with Estimated Bills in Your Analysis
Bobby, this is always tricky. My approach depends on the percentage and pattern of estimates. If it's random estimates scattered throughout, I usually keep them but note the limitation in my report. If there are consecutive months of estimates, I'll often exclude that period from trend analysis but include it in total cost calculations. The key is transparency with your client about how estimates affect your findings.
I've had good luck using degree day analysis to validate estimated bills when weather data is available. If the estimates seem reasonable based on heating/cooling patterns, I'll include them with a note. If they're way off, I exclude them and adjust my analysis period accordingly. Also always check if there was a true-up adjustment when actual readings resumed.
For demand analysis specifically, I create separate calculations with and without estimated periods. Estimated demand readings are often just copied from previous months and don't reflect actual peak usage. This can really skew your ratchet calculations and rate schedule recommendations. Document both scenarios in your report.
Had a similar situation with Xcel Energy last year. What I did was request interval data from the utility for the periods with estimated bills. Sometimes they have the actual usage data but just estimated the bills for operational reasons. Worth asking - got us the real numbers and made the audit much more accurate.
Larry makes an excellent point about interval data. Also check if the customer has their own metering or energy management system that recorded actual usage during estimated periods. We've been able to reconcile several audits this way. Sometimes the customer data is more accurate than the utility estimates.
All great suggestions. One thing to add - always calculate the potential impact of the estimates on your final savings numbers. If estimates could materially affect your conclusions, consider expanding your analysis period to include more actual readings, even if it means going back further or waiting for additional bills. Client confidence in your numbers is worth the extra time.