A commercial utility bill is read in three layers: the header (who the account is and what tariff it's on), the line items (every charge, grouped into energy, demand, fixed, riders, and taxes), and the totals (what the charges sum to and whether they tie out). A residential bill has about four lines. A commercial and industrial (C&I) bill on a time-of-use demand tariff can have thirty to fifty — and the single largest charge is often not energy at all, but demand.

This guide walks through each layer, defines every charge type you'll see, and explains the billing artifacts — estimated reads, mid-cycle rate changes, ratchets — that make C&I bills harder to read than they look.

#The three layers of a commercial utility bill

Every C&I electricity bill, regardless of utility, organizes into the same three layers even though the layout differs on every one:

  • Header — account identity and the tariff that governs every charge below it.
  • Line items — the charges themselves, each with a quantity, a rate, and a subtotal.
  • Totals and reconciliation — the sum of the line items, the printed total, and any rounding or off-bill adjustment between them.

Read in that order. The header tells you which rules apply; the line items are instances of those rules; the totals tell you whether the bill is internally consistent.

#Layer 1: the header

The header is the part people skim, and it's the part that gives every line item its meaning. A $14,000 demand charge is reasonable or absurd depending entirely on what the header says the account is. Key fields:

  • Account number and service address — identity.
  • Meter ID — which meter the reads came from; matters when a site has more than one.
  • Service period — the start and end dates the bill covers. Charges are computed over this window, and a window that spans a rate change gets pro-rated (see below).
  • Utility and tariff schedule code — e.g. PG&E E-19, Con Edison SC-9, Duke GS-T. This code is the single most important field on the bill: it determines the TOU windows, the demand structure, the riders, and the rates.
  • Rate class / customer class — commercial, industrial, agricultural; this and the tariff code together fix which charge types should appear.

Without the header, the line items are just numbers. With it, you know what each number should be.

#Layer 2: the line items

This is where the bill's complexity lives. C&I charges fall into five groups.

#Energy charges ($/kWh)

What you pay for electricity actually consumed, in dollars per kilowatt-hour. On a time-of-use (TOU) tariff, energy is split by period — and seasonally:

  • Summer on-peak, summer mid-peak, summer off-peak
  • Winter on-peak, winter mid-peak, winter off-peak

Each period has its own rate, with on-peak the most expensive. A single C&I bill can carry six or more energy lines before any other charge type appears.

#Demand charges ($/kW)

What you pay for the rate of consumption — the highest power draw during the billing period, in dollars per kilowatt. Demand charges are frequently the largest single component of a C&I bill, and the one residential customers never see. Common variants:

  • Coincident demand — your peak measured during the utility's system-peak window.
  • Non-coincident demand — your own peak, whenever it occurred.
  • Ratchet adjustment — a demand floor. Many tariffs bill demand as a percentage of your highest peak over the trailing 11–12 months, so a single high month sets an elevated floor that follows you for a year (covered below).
  • Power-factor adjustment — a surcharge or credit based on how efficiently the load uses power.

#Fixed charges

Charges that don't move with usage — a monthly service or customer charge, metering charges, minimum bills. Solar and storage don't offset these, which is why they matter when reconstructing a savings baseline.

#Riders, adjustments, and pass-throughs

Utility-specific surcharges layered on top of the base rate:

  • Transmission and distribution service charges
  • Public benefit / public purpose charges
  • Franchise fees
  • Fuel or purchased-power adjustments

The same charge often has different names across utilities — "public benefit charge," "public purpose programs," "PBC" — which is one reason bills from different utilities are hard to compare line by line.

#Taxes

State, county, and city taxes, plus utility users' taxes where they apply. Usually the last lines before the total.

#Layer 3: totals and reconciliation

The bill sums the line items into a printed total. The two should match — but utility bills round at the line-item level and again at the subtotal level, so there's often a small rounding delta. Occasionally a credit is applied off-bill (a demand-response payment, a prior-period correction) that reduces the total without printing as a line.

When you read a bill, the final check is whether the line items you can see sum to the total that's printed. If they don't, either there's a rounding adjustment, an off-bill credit, or a line you misread. A bill where the parts don't tie out to the total is a bill you can't yet trust.

#The billing artifacts that trip people up

Three normal utility behaviors make a bill look wrong when it isn't. Misreading them is the most common source of error in any manual bill analysis.

Estimated reads. When the meter isn't read this cycle, the utility estimates usage, prints an "EST" flag, and bills the estimate. The next cycle trues up against the actual read — often with a negative adjustment that makes one month look anomalously low. Two consecutive bills only make sense read together.

Mid-cycle rate changes. When a utility's new tariff takes effect on a date inside your service period, the bill pro-rates: the same charge type prints twice, once at the old rate, once at the new, split by the days each covered. Two "Energy Charge" lines on one bill usually means a rate change, not a duplicate.

Demand ratchets. A ratchet bills your demand as a percentage (often 80–100%) of your highest peak over the trailing year — not just this month's peak. So a single hot afternoon in July can set a demand floor you keep paying against through the following spring, even in months your actual peak was far lower. If a demand charge looks too high for the month's usage, a ratchet from an earlier month is the usual reason.

#How to read a commercial utility bill: the order of operations

Putting it together, read a C&I bill in this sequence:

  1. Header first. Note the tariff code, rate class, and service period — they define what everything else should be.
  2. Group the line items into energy, demand, fixed, riders, taxes. Expect multiple TOU energy lines and one or more demand lines.
  3. Check for artifacts. Scan for an estimated-read flag, two lines of the same charge type (rate change), and a demand charge that looks disproportionate (possible ratchet).
  4. Tie out the total. Sum the line items and compare to the printed total; account for any rounding or off-bill adjustment.
  5. Map to tariff components if you're using the data downstream — match each line to its role in the tariff so it can feed a model, proposal, or audit.

Step 5 is where reading a bill becomes using one — and where, at any scale beyond a handful of bills, doing it by hand stops being viable — and generic OCR isn't enough either.

#What Tariform does

Reading one bill is an afternoon's work. Reading hundreds — across multiple utilities, every month, into structured data you can put in a model — is a different problem.

Extract turns utility-bill PDFs into line-itemized, tariff-aware, source-traceable data: every charge mapped to its role in the source tariff, estimated reads and mid-cycle changes handled as known patterns, every bill reconciled against its printed total or flagged for review, every value pointing back to its source PDF line. It's built for project finance analysts, energy consultants, and C&I solar sales engineers who turn bills into pro-formas, deliverables, and proposals.

Book a demo — twenty minutes, a real bill, you see the output. Or start a free trial and upload one yourself.

#FAQ

Which date should I use for analysis — the bill date, the due date, or the service period?
The service period. Charges are computed over the start–end window the bill covers, so that's the range to align usage, savings, or tariff lookups against. The bill date and due date are administrative — useful for cash flow, not for matching a charge to the month it was incurred.
My account has more than one meter — how does that show up?
A multi-meter account either prints a separate line-item block per meter or aggregates them under one total, depending on the utility; the meter ID in the header is what disambiguates them. When you reconcile, check whether demand is summed across meters or billed per meter — tariffs differ, and it changes the demand charge materially.
Where do I find the actual rates behind my tariff code?
Not on the bill. The bill shows the rates applied; the governing schedule — every TOU window, demand structure, rider, and ratchet clause — is the published tariff sheet on the utility's or state regulator's (PUC/PSC) website, keyed by that schedule code. The code on the bill is the pointer; the tariff document is the source of truth.
How do I compare bills from two different utilities?
Match on charge role, not charge name. The same component — a distribution charge, a public-benefit rider — is named differently across utilities, so a name-by-name comparison misleads. Group each bill into energy, demand, fixed, riders, and taxes, then compare the roles; that's the only basis on which two utilities' bills are comparable.
What are the kVA or kVAR lines on my bill?
They relate to power factor. kW is real power (what does work); kVA is apparent power (what the utility must supply); kVAR is the reactive portion between them. Motor-heavy loads draw more apparent than real power, and a tariff may bill demand in kVA or add a power-factor charge to recover the difference — which is why a low-power-factor site can show a demand line larger than its kW alone would suggest.
How do I turn a stack of these bills into data I can model?
Read each in the order above, map every line to its tariff role, reconcile to the printed total, and record where each value came from. That's an afternoon for one bill and unviable by hand for hundreds — which is what Extract automates, turning bill PDFs into line-itemized, tariff-aware, source-traceable data.