Outage Operations: Monitoring and Planning for a Weather-Driven Outage Event

Overview

This tutorial walks through how to use Outage Operations across the full lifecycle of a weather event from routine monitoring through active event response and after-action review. Steps use the Predict tier as the base workflow. Where Predict Plus or Restore capabilities extend a step, those are called out inline.

What you will do in this tutorial:

• Check the five-day outage forecast as part of your daily monitoring routine

• Identify high-risk days and service centers as a weather event develops

• Drill into the forecast drivers and assess confidence as the event approaches

• Use the forecast to support crew pre-staging, mutual aid, and customer communication decisions

• Track conditions on the day of the event

• Review forecast performance after the event

Step 1: Open the Dashboard and review the five-day forecast

Open Outage Operations and select Dashboard from the left navigation bar.

The Dashboard opens to the most recent forecast run. The row of day cards at the top shows the territory-wide predicted outage count for each of the next five days. The color coding surfaces days with elevated forecast counts immediately.

What to look for:

• Are any of the next five days showing an elevated count relative to your normal baseline?

• Is the high-count day isolated or does it span multiple consecutive days?

• Does the severity percentile (shown in the district table) indicate this is a routine event or an outlier relative to your storm history?

If all five days show low counts with no color escalation, your territory is not under a known outage threat for that window. No further action is required until the next forecast run.

If one or more days shows an elevated count, continue to Step 2.

Step 2: Identify where the impact is concentrated

Select the highest-count day card to filter the district map to that day.

Scan the district table for service centers with the highest forecasted outage counts. The color coding matches the map, so you can cross-reference geographic clusters with the tabular data simultaneously.

What to look for:

• Are high-count service centers geographically clustered, or distributed across the territory? A cluster suggests a localized meteorological driver. Distribution across the territory suggests a broad-area event.

• Which service centers are consistently appearing at the top of the table across multiple high-count days?

Select any service center row in the district table to navigate to that service center's detail view on the Map.

Step 3: Examine the meteorological drivers on the Map

Select Map from the left navigation bar.

In the top navigation bar, confirm Outage Forecast is selected as the active layer under the Outages group. The map shows service centers color-coded by forecasted outage count for the day shown on the timeline.

To understand what is driving the forecast, switch to a weather variable. In the Risk and Weather Layers selector, choose a relevant variable depending on the expected hazard type. Use the timeline to step through the forecast hours and observe when and where conditions peak.

Toggle back to Outage Forecast to compare the spatial pattern of predicted outages against the meteorological driver you just reviewed. The highest outage counts should align with the areas of most severe forecast conditions.

What to look for:

• Does the spatial pattern of the outage forecast match the meteorological pattern? If so, the forecast is behaving as expected.

• What time of day do conditions peak? This affects the timing of crew deployment decisions.

• Are NWS alert polygons active over your highest-risk service centers? Toggle the NWS alerts layer on via the Layer Manager to overlay official watches and warnings.

Select any weather station marker on the map to compare current observed conditions against the forecast at that location. As the event approaches, this comparison becomes a useful confidence check (observed conditions tracking toward the forecast supports higher confidence in the predicted outage counts).

Step 4: Assess forecast confidence using the evolution chart

Return to the Dashboard and review the Forecast Evolution chart.

This chart shows how the predicted outage count for each day has changed across successive forecast runs. A prediction that has been stable across the last several runs indicates higher meteorological confidence. A prediction that has shifted significantly between runs (either in magnitude or in which days carry the highest counts) indicates that the underlying weather forecast is still evolving and that planning decisions may warrant a wider range of scenarios.

What to look for:

• Has the forecast been consistent for three or more consecutive runs? If so, you can plan with reasonable confidence against the current prediction.

• Is the forecast magnitude trending upward or downward as the event approaches? Upward revision for a multi-day event is common as meteorological guidance sharpens.

• Are high counts shifting between days across runs? This may indicate uncertainty in event timing rather than overall severity.

Use this context to inform how aggressively you pre-position resources. A stable, high-count forecast several days out warrants earlier action than a volatile forecast of similar magnitude.

Coming Soon: Step 5: Drill into service center detail

On the Map, select a high-priority service center to open the District Detail Panel.

Select the Outages tab. Review the predicted outage count and severity group for each day in the forecast window.

Select the Forecast tab to review the hourly weather forecast for this service center. This shows you the specific conditions driving the outage prediction for that location and is useful for briefing field crews or preparing stakeholder communications with location-specific detail.

Repeat this step for each high-priority service center identified in Step 2.

Coming Soon: Step 6: Support resource and mutual aid decisions

Using the information from Steps 2 through 5, the forecast now supports the following planning actions.

Crew pre-staging: Match crew type and location to the geographic concentration of high-count service centers identified in Step 2. Earlier action is warranted where the forecast evolution in Step 4 shows a stable, high-confidence prediction.

Mutual aid coordination: Share the outage count forecast and, where applicable, the damage category breakdown and crew count estimates with neighboring utilities early enough for them to commit resources before the event arrives. Outage Operations forecasts are available up to five days in advance specifically to support this lead time.

Material procurement: For events with a high proportion of equipment-related damage in the Predict Plus forecast, use the damage category counts to estimate quantities of poles, transformers, and other restoration materials needed. Stage these in or near the highest-risk service centers where logistics allow.

Emergency Operations Center (EOC) activation decision: Use the territory-wide outage count and severity group to assess whether the event meets your thresholds for standing up your EOC.

Step 7: Prepare customer and stakeholder communications

The outage count forecast and severity context from Outage Operations directly support two external communication needs.

Customer outreach: Use the territory-wide count and the list of highest-count service centers to identify which customers are in the projected impact zone. Proactive outreach before the event reduces inbound call volume during restoration.

Internal briefings and reports: Share the outage forecast and district-level breakdown for distribution to operations leadership, EOC staff, or regulatory contacts. Include forecast evolution context in briefing materials to show how the prediction has developed over time.

Step 8: Monitor conditions on the day of the event

On the day the event is expected to arrive, return to the Map and use it for situational awareness alongside active restoration operations.

Track actual versus forecast conditions: Select weather stations in the affected service centers to compare observed wind speed, gust, and other conditions against the forecast values for that hour. Conditions tracking above forecast values may indicate that actual outage counts will exceed predictions.

Monitor incoming storm position: Track storm position in real time and use this alongside the weather forecast to refine the timing of crew deployment and EOC briefing cycles.

Update resource deployment as needed: If observed conditions or actual outage counts are diverging significantly from the forecast, use the service center detail views to re-examine the remaining forecast hours and adjust crew positioning accordingly.

The forecast continues to update twice daily at 00Z and 12Z throughout the event, so the on-screen forecast always reflects the latest available meteorological guidance even as the event is in progress.

Step 9: Conduct an after-action review

After the event has passed, use Outage Operations to review how the forecast performed against the actual event.

What to examine:

• Which service centers were accurately forecast? Which were significantly over- or under-predicted?

• How did forecast accuracy change as the event approached across successive runs?

• Were there service centers where the damage category breakdown (Predict Plus) aligned with the actual repair mix encountered by crews?

Document these findings for two purposes. First, they inform future pre-event planning by building institutional knowledge of how the model performs for specific event types and geographic areas in your territory. Second, sharing actual outage data from the event with Technosylva supports model retraining, which progressively improves forecast accuracy over time.