Definition
In the "Building Energy Benchmarking and Sub-metering" scenario, sensors and business terminals often connect via cellular IoT platforms; coverage quality, latency, security, and the full lifecycle of the SIM card must all be considered.
Technical Features
It is recommended to differentiate telemetry, alarms, configuration, and OTA traffic; configure caching and retransmission in weak network environments; include TLS, certificates, and keys in the change and rollback strategy.
Use Cases
Applicable to distributed sites and remote operations. SimtoLink can provide combined capabilities like connection management, data pool, and directed/private networks, supporting large-scale replication.
Key points for cooperation with IoT SIM cards
- Ledger: ICCID / IMEI bound to device number, installation location, and person in charge.
- Network: APN, whitelist, SIM-device binding, and regional/roaming policies.
- Operations & Maintenance: online rate, abnormal traffic, silent SIMs, and alert closure in ITSM.
- Compliance: log desensitization, retention period, and cross-border/subcontracting agreements.
Submeter definitions, gaps, and the benchmark period
A building energy benchmark asks whether this building, under the same weather and occupancy, has left its own baseline. It is not a raw comparison of two buildings’ meter readings. Submetering separates HVAC, lighting, plug loads, special power, and water services. The baseline is drawn by period — a day, a month, or a cooling season — and by the normalizing conditions. The IoT SIM connects the submeter or the collection gateway and carries timestamped electricity, cooling, and flow to the energy platform. The technical point is the definition: the same submeter category in different buildings must mean the same kind of load, and a missing interval must be marked empty, not written as zero. Zero is counted as “very efficient” and pulls the baseline crooked.
When cellular aggregates many buildings, each gateway needs its own identity, a clock that matches the platform, and a reporting period aligned with the benchmark window. If the benchmark uses hourly energy and the gateway uploads one total at dawn, the platform cannot see peak and valley, or which hour was missing. A meter replacement must keep the closing reading and the multiplier. A new meter that starts at zero cannot be subtracted directly from the old meter. Restricted access should reach only the energy platform, so other office traffic does not mix onto the same SIM and sit inside the “energy curve.” If a meter is removed during construction, pause the SIM or mark it as not metering. A long run of zeros will otherwise enter the baseline sample. If several buildings share one egress address, still distinguish them by device id. The source address alone cannot say which building it was.
A common error is to subtract submeters from the main meter, call the remainder “other,” and then treat that remainder as a measurement. A few minutes of timestamp difference between the main meter and the submeters produces negatives or spikes in the difference, which an efficiency report then writes up as abnormal waste. Another error is a peak-valley tariff window that does not match the meter’s freeze time, so off-peak energy is counted as peak. A gateway that is offline for several days and then catch-up-uploads in one burst is booked on the arrival day, so the baseline month loses some days and gains others. A dashboard that reads every point every second turns a light metering SIM into a high-volume SIM and adds nothing to the benchmark. Acceptance should take one full cooling month and check that the sum of submeters versus the main meter, the count of missing hours, and the multiplier-change record can all be traced to a SIM identity and a meter id. The tariff window is described under time-of-use tariff. The control system that holds the heating and cooling points is described under HVAC.
Baseline samples should drop days of fit-out, vacancy, and meter faults in advance. If those days are treated as normal use, the next season’s benchmark is systematically high or low. Keep missing hours as empty values and list the hour count in the monthly report. Do not average the hours on either side and present the result as measured. Multiplier changes, current-transformer changes, and changes to the submeter definition all need an effective date. Data before that date still uses the old definition in the report. Buildings in one portfolio can start and end the cooling season on different days. Do not force them onto one calendar month. On the day a gateway is replaced, prevent the old and new meters from both adding the same energy, and prevent a negative. A curve cited in a benchmark conclusion should click back to the SIM identity and the meter id, so an audit can tell whether a low-use hour was a real saving or a communications gap. A before-and-after comparison of an efficiency retrofit should cover the same kind of weather across a full meteorological cycle. Comparing one week before and one week after easily writes a weather change down as the retrofit effect.
Write the gateway’s clock source into the benchmark note. If the gateway’s own oscillator, the time pushed by the platform, and the meter’s internal freeze time each run alone, the same hour produces three energy figures. The benchmark accepts only the set whose source is named. The others are a check. If a submeter is moved and the load mix changes, split the baseline from the change date. Do not join the months before and after lighting was moved onto the plug meter into one falling curve. If a fresh-air fan and a chiller share one meter inside HVAC electricity, the report should not say both submeters have been measured. When cellular catch-up crosses a month boundary, energy belongs to the freeze month, not the month the catch-up arrived. The monthly report may be revised the next month, and the revision must keep the original value. When many buildings share a data pool, a firmware upgrade in one building should not push the others’ reports outside the benchmark window. Put the upgrade window in the monthly note so a communications gap is not explained as a drop in energy use.