Reading a household electricity meter on a fixed schedule: an observation protocol

Goal

Produce a series of raw meter readings that a second household can copy in form, so that daily energy differences can be compared and explained by the recorded conditions rather than guessed.

Prerequisites

Visual access to the meter display, a clock, a notebook or a text file, and agreement in the household on who reads and when. The protocol reads the display only: no cover other than a viewing door is opened, and no seal, fuse, wire or terminal is touched. If the display cannot be seen without doing any of that, the protocol does not apply.

Steps

1. Identify the meter once: serial number as printed, type (mechanical drums or electronic display), the unit printed next to the reading (usually kWh; NIST SP 811 lists the kilowatt hour with the factor 3.6 MJ per kW·h), the number of decimals shown, and how many registers exist (tariff registers, an export register on a house with solar panels). Electronic displays often cycle through several values, so record the label shown next to each value and state which register the log follows.
2. Fix a reading time (for example 07:00 local time) and a daily cadence. Read within a tolerance you write down, such as plus or minus 15 minutes; if a reading is late, log the actual time, never the planned one.
3. Write down the raw value exactly as displayed, including leading zeros and decimals, with date and time. Do not round and do not subtract yet.
4. In a conditions column, note what plausibly matters that day: number of people at home, heating or cooling on, laundry or dishwasher runs, guests, absence, a new appliance.
5. Compute daily differences later in a separate column or file and keep the raw log untouched. Mark a reading you doubt (a drum sitting between digits, a display caught mid-cycle) instead of correcting it.
6. After a window of at least 14 days, report the raw table, the smallest and largest daily difference, the register followed, the displayed resolution, and the conditions log.

Expected result

A table another household can reproduce in structure; differences between two logs can be traced to conditions, resolution or register rather than to method.

Limits and test basis

Proposed protocol; no consumption figure is claimed. The display resolution bounds the smallest daily change that can be seen, and a cumulative meter cannot attribute energy to individual appliances. Nothing here concerns wiring, tariffs or safety; the only action is reading numbers off a display.