Everything on this site rests on a document most readers have never opened. The tariff is public, it is authoritative, and it settles every question that a bill leaves ambiguous.
It is also written for a regulator rather than for you, which is the actual obstacle. The style is precise and unwelcoming, cross-references run between sheets, and the single provision that decides your bill may sit three documents away from the rate table. None of that makes it hard to read. It makes it hard to read casually, which is a different problem with a straightforward solution: know what you are looking for before you start.
Where the document lives
Two sources, and the distinction matters.
The utility's own tariff library is convenient and usually current. The state commission's filing is authoritative. Approved tariffs are public records, and where the two disagree, the filed version governs. For anything you intend to act on, use the commission's copy.
You also need the version in force for the period you are examining. Auditing last summer's bills against this year's tariff will produce discrepancies that are entirely your own doing.
The structure
Tariff books are organized in layers, and knowing the layers saves a great deal of searching.
General rules and regulations apply to every customer: definitions, metering, billing practice, disputes, service conditions. This is where the meaning of terms used throughout the rate schedules is fixed, and it is skipped by nearly everyone.
Rate schedules are the individual tariffs — one per customer class and service type. Each has applicability, character of service, rates, and special provisions.
Riders and adjustment clauses are separate documents applying across schedules, each with its own basis and its own update cycle: riders and surcharges.
Contract forms and service agreements cover special arrangements, standby service and interconnection.
The trap is that a single schedule read in isolation is incomplete. The demand definition may sit in the general rules; the effective demand price is the schedule rate plus every applicable rider; the ratchet may be defined once and referenced everywhere.
The six things to extract
Reading a tariff end to end is not the objective. Extracting six specific provisions is.
1. Applicability. Who may take service under this schedule, on what criteria, measured over what period. The measurement basis matters as much as the threshold: a limit tested over twelve months is a different constraint from one tested monthly.
2. Determination of demand. The interval length, whether it is fixed-block or rolling, whether demand is measured across all hours or inside a window, and whether more than one demand determinant is billed. See the demand interval and facility, on-peak and billing demand.
3. The ratchet. Its existence, percentage, look-back period and which months qualify. This single clause can be the largest difference between two schedules: ratchet clauses.
4. Power factor. Whether it is penalized, whether demand is billed in kVA, or whether it is ignored entirely. Three different treatments, three different investment cases — and a capacitor bank that pays back handsomely under one of them and not at all under another.
5. Season and period definitions. When summer starts and ends, when the on-peak window runs, which days count, and which holidays are excluded by name. Control strategies are built on these and are wrong if they are wrong.
6. The applicable riders, each with its unit. Nothing can be priced without them.
With those six, plus the rate tables, you can reproduce any bill from your own interval data — which is the test of whether you have read it correctly.
Reading conventions worth knowing
Definitions are binding and are not always intuitive. "Month" may mean a billing period rather than a calendar month. "Demand" without qualification usually means the measured maximum, but the qualified terms elsewhere may differ.
"Shall" is mandatory and "may" is not. A schedule that says service shall be taken under it above a stated demand level is describing an automatic reassignment, not an option.
Sheet numbers, revision numbers and effective dates matter. Tariff sheets are revised individually. A sheet marked as a revision supersedes a specific earlier sheet, and the effective date tells you which periods it governs.
Cross-references are load-bearing. A phrase like "as defined in the General Rules" is not decoration; the definition it points at is the one that applies.
Special provisions carry the exceptions. Most schedules end with a section of special provisions, and it is where minimum bills, contract requirements, seasonal exceptions and the awkward cases live. A schedule read down to the rate table and no further has skipped the part that most often explains a bill.
The test of a correct reading
You have read the tariff correctly when your model of it reproduces a bill you have already received.
Take one billing period, take your interval data, apply every provision you have extracted, and compare the result with the actual statement line by line. If they agree, your understanding is right and any comparison you build on it is trustworthy. If they do not, the difference is almost always a clause you have not found — a ratchet, an adjustment, a multiplier — rather than a utility error.
That reconciliation is the foundation of every serious tariff exercise: it is what makes how to choose a rate schedule a calculation rather than an opinion, and it is what turns a bill audit from a review into a proof.
It is also, unglamorously, the reason this article exists. Every recommendation on this site is conditional on a document only you can read, for a site only you can see. Reading it is not a formality before the interesting work. It is the work.
