Advanced alarms rules and analysis

Advanced alarms turn SmartgridX telemetry into an operational workflow. Instead of only reporting a manufacturer fault, the alerting engine evaluates whether an installation is still communicating, following control instructions, respecting grid limits, producing as expected, and delivering the intended result.

From signal to action

This separation is important: a missing signal is not treated as proof that an operating limit was violated, and a temporary scheduler wait is not shown as a customer alarm.

Rule catalogue

Grid safety

Compares the site's quarter-hour average grid power with the maximum import and export limits configured on the controller. Positive grid power represents import; negative grid power represents export.

The rule fires when either:

• grid import exceeds the configured maximum consumption; or
• grid export exceeds the configured maximum injection.

The analysis chart shows the measured grid power together with both limits, so the direction and size of a breach are visible immediately. The rule is offered only for sites where these limits are configured.

PV status

Compares solar inverters at the same site. It fires when one or more inverters produce almost nothing while their peers clearly produce. This makes it useful for finding an offline inverter or string without relying on a weather forecast.

• If all inverters are inactive, for example at night or during very low irradiance, the rule does not infer a failure.
• A site with only one inverter cannot use peer comparison to reach the same conclusion.
• A missing inverter signal remains distinguishable from a confirmed production anomaly.

External signal

Supervises external control signals on the site and all descendant devices.

1. Signal liveness: fires when no new external-signal sample has arrived for more than 1 hour. The evaluator looks back 24 hours to report the most recent activity.
2. Setpoint tracking: when the signal contains a power setpoint, SmartgridX compares it with the asset's actual power over the last 2 hours. A point deviates when the difference is greater than the larger of 3 kW or 10% of the setpoint. The rule fires when more than 30% of the aligned points deviate and the latest point still deviates.

Setpoints and measured power are aligned by carrying the most recent observation forward to the setpoint timestamp. This avoids comparing values from unrelated moments.

Battery status

Detects a state-of-charge signal that appears to be stuck. By default, the highest and lowest SoC values over 48 hours must differ by at least 1 percentage point.

To prevent false alarms on a newly connected battery, the evaluator requires at least two measurements and data covering at least half of the 48-hour window. Otherwise the result is No data, not Firing. On a site with multiple batteries, one stuck battery is sufficient to open the alarm, and the analysis view identifies the affected device.

Savings

Checks whether estimated EMS savings continue to grow over a minimum period of 7 days. The rule is offered only when SmartgridX has savings data and at least one asset is configured for cost optimisation.

The analysis can separate the contribution of batteries, PV, EV charging, and HVAC or switched loads, and it surfaces invalid or incomplete source data. Savings alarms are intended for operational analysis in the application and are not included in customer alarm-email bundles.

Custom threshold

Creates a rule from a telemetry signal available at the selected site or one of its devices. Choose:

• the measurement;
• above or below;
• the numeric threshold; and
• how long the condition must persist.

Available measurements are discovered from the installation and can include grid power, production, battery power, EV charging power, base load, state of charge, planning values, setpoints, limits, and prices. Always confirm the displayed unit before saving the rule.

Understand the status

For firing alarms, SmartgridX also distinguishes a recent outage from a signal with no recent healthy activity. The last known healthy timestamp helps separate a fresh failure from an installation that was never fully commissioned.

Use the analysis view

Open Analysis from a rule card before acting on an alarm. Depending on the rule, the view can show:

• the measured series and its threshold or import/export limits;
• actual power next to the requested setpoint;
• one chart per affected inverter or battery;
• strategy overlays and categorical control-state bands;
• shaded breach intervals;
• followed, constrained, and violated setpoint samples; and
• the last healthy activity and duration of the current violation.

The small chart on a rule card is intentionally a summary. Use the full analysis view to verify timing, units, and the relationship between signals.

Create and manage rules

1. Open Advanced alarms for the correct SmartgridX organisation.
2. Select New alarm rule and choose a rule type.
3. Select an installation. For automatic rules, the selector shows only sites with the required assets, signals, or controller settings.
4. Review the fixed rule definition or configure the threshold fields.
5. Add an optional recognisable name and save the rule.
6. Open Analysis after the first evaluation to verify that the expected signals were selected.
7. Add notification recipients only after the rule has been validated.

Rules can be enabled or disabled without deleting their configuration. Operators can also edit or remove a rule, filter the overview by type and status, group results by installation, and manually request a recheck.

Configure a portfolio efficiently

Bulk configuration can create applicable automatic rules across multiple sites. SmartgridX first checks each site's eligibility, so a battery rule is not created for a site without a detected battery, and a grid-safety rule is not created without grid limits.

Use individual configuration for custom thresholds and 7C compliance: both require an explicit operational choice rather than generic portfolio-wide defaults. Account or subscription limits can restrict how many rules are active.

Notifications

Email notifications are bundled per recipient and organisation. A recipient can choose a minimum interval of 1 hour, 3 hours, 6 hours, or 24 hours. A bundle can contain both newly opened alarms and resolved alarms.

To avoid notification floods:

• email is never sent more frequently than once per hour to the same recipient;
• repeated evaluations of the same open condition are deduplicated;
• a short warm-up period after an alerting-service restart allows all rules to be evaluated before notifications resume; and
• configured SMS uses the same cadence and reports new alarms only.

Troubleshooting checklist