Top consumers
Meters ranked by consumption, highest first
Live load curve
Demand (kVA) — last 90 readings, updates every 20s
Power factor
Target ≥ 0.90
Demand utilisation
Against rated capacity, where one is set
AI cost forecast
Next 7 days projected from recent trend — –
Consumption
kWh consumed — last 30 days
Consumption pattern
Average load by hour of day — shaded by tariff band (peak / normal / off-peak)
Peak vs. off-peak consumption
Power quality & harmonics
Frequency and distortion — –
Harmonic spectrum — odd orders as % of fundamental
Three-phase voltage
Nominal 415 V
Three-phase current
Rated –
Transformer step-down loss
Mains feeder (HT) input vs main incomer (LT) output, by location
About this page
Deeper, per-parameter graphs for – — each with its own hourly/daily/weekly/monthly view, a live snapshot, and a plain-language explanation of what the parameter means and why it matters for this equipment.
Voltage
Electrical potential difference between two points, measured in volts (V) — the "pressure" pushing current through a circuit. Indian LT panels typically run at 415V line-to-line; HT feeders at 11kV or higher. Voltage should stay within a narrow band of nominal — sustained deviation can trip equipment or shorten motor life.
Live — nominal 415 V
Current
The rate of flow of electric charge, measured in amperes (A). Current rises and falls directly with load — as machines switch on, current increases. Every conductor and breaker has a maximum current rating; running near or above it for long periods causes overheating and is a common cause of cable and panel failure.
Live — rated –
Power factor
The ratio of real power (kW) doing useful work to apparent power (kVA) actually drawn from the supply, from 0 to 1. A PF of 1.0 means all supplied power is used productively; a lower PF — common with motors, transformers and other inductive loads — means the utility must supply more current than the useful work requires. Indian tariffs typically penalize average PF below 0.90 and reward PF above 0.95.
Meter-wise power factor
All connected meters compared side by side, lowest (most at-risk) first
Power factor pattern by meter
A quick shape-of-the-pattern view for every meter — daily (last 14 days), weekly (last 10 weeks) and yearly (last 12 months).
Apparent power (kVA)
The total power the supply must deliver — real (working) power plus reactive (non-working) power — measured in kilovolt-amperes. kVA = kW ÷ power factor. Sanctioned/contract demand and transformer or panel capacity are all rated in kVA, since the supply infrastructure has to carry the full apparent power regardless of how much of it does useful work.
Active energy (kWh)
The energy actually consumed and converted into useful work — motion, heat, light — measured in kilowatt-hours. This is what appears on the electricity bill as energy charges: 1 kWh is 1 kW of power sustained for 1 hour. Hourly shows consumption in each hour of today; daily/weekly/monthly show totals for that period, not an average.
Reactive energy (kVArh)
Energy that oscillates between the supply and inductive loads — motors, transformers, ballasts — without doing useful work, measured in kilovolt-ampere-reactive-hours. It isn't billed directly like kWh, but a high kVArh relative to kWh (i.e. low power factor) usually triggers a power-factor penalty and increases losses in cables and transformers. Like kWh, periods other than hourly show totals, not averages.
Load factor
How evenly load is drawn, rather than spiking and idling — average demand as a percentage of peak demand over the period. A plant running steadily near its peak all day has a high load factor; one that spikes briefly and idles the rest of the time has a low one, even with the same total energy bill. Utilities often reward a high load factor (MSEDCL's HT tariff, for instance, pays an incentive above 75%) since it means better use of the capacity they've built for you. Shown as an average of each day's load factor, not hourly — within a single hour average and peak demand are nearly the same thing, so an hourly figure wouldn't be meaningful.
Custom date range report
Generate a consumption & cost report for any period, per meter or aggregated.
Daily grid energy & forecast
Actual kWh per day, with the 14-day forecast
Health score
How the score is built
Weekly pattern
Average grid kWh by day of week (last 8 weeks)
Daily load profile
Average grid demand by hour
Savings opportunities & risks
Calculated from your data and tariff — each shows its working.
AI report & recommendations
Written by the Ollama agent from the figures above.
Ask the agent
It looks up the data it needs (meters, history, alerts, live readings) before answering.
Meter breakdown
Selected period
| Meter | kWh | Share | vs prev | Peak kVA | PF | LF |
|---|
Anomaly detection (baseline agent)
Nightly statistical baselines per meter: anomalies, trends and ratchet risks. These are fed to the AI agent too.
Alert log
Demand, power factor and fluctuation events across all meters.
WhatsApp / email delivery shown below is simulated for this prototype — connect a messaging gateway in Settings → Notification channels to send real alerts.
| Time | Meter | Type | Severity | Message | Channels |
|---|
Automatic cut-off control
Rules watch voltage, load and demand, then publish a cut-off command to your MQTT broker. A relay, PLC or gateway subscribed to that topic opens the breaker.
Rules
Each rule watches one meter. It trips after the value stays past the trip level for the hold time, and restores after it stays on the safe side of the restore level.
| On | Rule | Meter | Trip when | Restore | Live value | Status | Actions |
|---|
Action log
Every command sent or attempted, plus arm/disarm and rule changes.
| Time | Event | Rule | Meter | Value | Topic / payload | Result | By |
|---|
PNG, JPG or WebP. A wide logo (about 3:1) on a white or transparent background looks best. Shown in the sidebar, the login page and on PDF reports.
Untick anything that doesn't apply to your metering setup — disabled checks stop alerting and, where they have one, their dashboard section hides.
Untick anything that isn't part of your billing structure — the field greys out and is excluded from cost calculations.
Click an hour to cycle it between Normal, Peak and Off-peak — defines when the peak/off-peak rates above apply.
These are placeholders for this prototype — wire in a WhatsApp Business API / SMTP gateway to make delivery real.
Simulated data runs out of the box with no setup. Switch to MQTT once a field gateway is publishing real readings to your broker.
The app subscribes to <prefix>/+/reading — publish each meter's readings as JSON to <prefix>/<meterId>/reading, e.g. dpinfotech/energy/m1/reading: {"demandKVA":312.5,"pf":0.93,"voltage":[414,416,415],"current":[145,146,144],"freq":50.01}. Any field can be omitted except demandKVA; harmonicsV/harmonicsI (6-element arrays) are optional too.
Add every meter that reports into this tenant. Protocol reflects how the field gateway polls the device. "Feeds from" links a Main Incoming meter to its upstream Mains Feeder (HT) meter to enable transformer step-down loss tracking. Set "Capacity?" to No for a meter with no rated kVA — it's tracked as energy-only, with no utilisation gauge or demand alert. The Meter ID column shows each meter's MQTT topic ID — click it to copy the full topic your gateway or Node-RED should publish to.