Sunstead
Sign in
Read the data
Find your home’s overnight energy baseline
The house is quiet, but the energy graph never reaches zero. Before looking for a fault, measure what an ordinary quiet night actually includes.
Choose a window you can repeat
Pick a period such as midnight to 5 a.m. and use the same local-time window on several nights. Record whether someone was awake, heating or cooling was running, an EV was charging, or an appliance had a delayed start. The goal is an ordinary baseline for your home, not the smallest number you can produce once.
Check the source’s measurement boundary. An inverter may cover only the circuits it serves. A charge-controller output is not a whole-home load measurement. Label the result “monitored circuits overnight” if that is what you have; a precise name is more useful than an overbroad total.
Calculate energy, then an average
Three fictional nights
Over the same five-hour window, suppose complete readings show 1.10, 1.35 and 1.20 kWh. Dividing by five gives average loads of 220, 270 and 240 W. A reasonable initial observation is that the measured overnight average was 220–270 W across those nights.
That range is not a guaranteed floor. One warm night, a guest or a pump cycle could change it. Retain the individual results alongside their notes instead of immediately reducing them to one permanent number.
If you only have snapshots, you cannot confidently derive the whole interval’s energy. Use a reported energy total or enough valid time-series observations to support the calculation. The data-coverage guide explains what to do with missing intervals.
Read the shape before naming the appliance
A roughly level trace suggests a persistent combined load. Repeated rises and falls suggest cycling activity, but the graph alone does not identify a refrigerator, pump or heater. Several appliances can overlap and resemble one larger device.
List known always-on items first: networking, security, refrigeration, standby electronics and equipment required by the property. Then add timed or thermostatic loads. Compare their normal schedules with the trace. If you investigate one item, change only an ordinary user control that is safe to change, and note the exact time. Keep essential equipment running.
Reject bad comparisons
A night with two hours of missing data should not be celebrated as an energy reduction. A midnight-to-5 a.m. total also cannot be compared directly with a six-hour window. Check clock settings, source changes and coverage before interpreting a smaller bar.
Weather can matter even when everyone was asleep: heating, cooling and refrigeration respond to conditions. Keep a short conditions note rather than forcing every night into one category. The Department of Energy’s storage overview explains the broader need to match stored supply with demand; your measured nighttime demand is one input to that planning.
Keep a one-page baseline record
Use columns for date, start/end time, monitored circuits, measured kWh, valid coverage, unusual loads and weather notes. Review a small set of comparable nights before choosing the next action. The record may show that the persistent load deserves attention, or that one intermittent activity dominates.
Carry the observed range into a battery plan, not just the best night. If the property changes—new occupants, a freezer, a pump schedule—start another comparison period. A baseline is a description of a particular operating pattern, not a judgment about whether the home is efficient.
Sources & further reading
Documentation checked October 10, 2026.
Your next step
Plan your first monitoring week