Load forecast in distribution network

1. Description of the Use Case

1.1. Name of the Use Case

IDArea /Domain(s)/Zone(s)Name of the Use Case
1Smart Grid / Distribution / Operation,PUC1

1.2. Version Management

Version No.DateName of author(s)ChangesApproval status
1.02020-12-04T00:00:00Luis Viguer,N/AApproved

1.3. Scope and Objectives of Use Case

ScopeBased on historical records, current load measures and weather forecast, it will be provided the total load forecast (consumption models) at a substation, primary or secondary, located at Kefalonia and Attica regions.
Objective(s)Provide input data to DSO in order to assess voltage levels and power flows at distribution network
Related business case(s)PUC12

1.4. Narrative of Use Case

Short description

Placed within the abovementioned BUCs, PUC1 describes how the forecast provider extracts, transforms and loads (ETL) the data needed by the DSO to assess voltage levels and power flows. It is composed by four Secondary Use Cases (scenarios):

  1. Get measurement form substation
  2. Get meteo forecast
  3. Generation of load forecast
  4. KPI calculation for load forecast

Complete description

This PUC makes reference to the forecasting modules focused on the operation of the DSO and the distribution network. In order to increase the performance of these modules, data requests to external systems will be sent. Finally, the results of the algorithms will be available to external actors/applications.

Type of considered forecasts: • Load Forecast: Forecast of total demand at a substation (primary or secondary).

The time-granularity/horizon of the forecast can be: • Very short-term: 1h (5 or 15 min steps), • Short-term: 24-48 h (15min-1hour step), • Medium-term: 48h-10days (1-hour step)

The PUC will deepen in the following steps that are necessary to forecast the total demand at a substation:

  1. Forecast Provider retrieves current load from substation.
  2. Forecast Provider asks for meteo forecast to Weather Forecast.
  3. Forecast Provider corrects wrong/missing data, using an algorithm for anomalies detection.
  4. Forecast Provider generates a mathematical model for load, using the historical data, and it applies the current load to the mathematical model, getting the load forecast.
  5. Forecast Provider sends the forecasts to DSO.
  6. DSO sends the forecasts to Coordinet platform (actors who need it for their activities).

Besides, KPIs related to accuracy of prediction will be calculated in order to get the efficacy of the forecast done.

Within the related business use cases, this primary use case, jointly with PUC2 (RES forecast in DN), focuses on the BUCs’ first step from DSO side:

More precisely, PUC1 is divided into four secondary use cases also referred to as scenarios:

• SUC 1.1 – Get measurement from substation: The Forecast Provider asks for current load to the Substation, which returns that measure.

• SUC 1.2 – Get meteo forecast: The Forecast Provider asks for the weather prediction to the Weather Forecast Provider, an external server which does not belong to Coordinet ecosystem. The request contains the moment of time to be considered, such as prediction for 1 hour by or 24 hours by. The response contains the weather forecasting for that moment.

• SUC1.3 – Generation of load forecast: After correcting wrong/missing data, the Forecast Provider generates a model of load for a period of time, using weather forecast and historical data related to load. Once this model is prepared, the Forecast Provider uses the current measurement to generate its prediction.

• SUC1.4 – KPI calculation for load forecast: The KPIs are calculated in order to determine if the prediction is good enough. This calculation is done considering the load prediction and the real load.

1.5. Key Performance Indicatiors (KPI)

IDNameDescriptionReference to mentioned use case objectives
KPI_24Accuracy of load forecast calculated 1 hour in advanceThis indicator measures the Mean Absolute Percentage Error (MAPE) of the load forecast in distribution system.N/A,

1.6. Use case conditions


1.7. Further information to the use case for classification/mapping

Relation to other use cases
Level of depth
Generic, regional or national relation
Nature of the use cases
Further keywords for classification
Smart Grid, Distribution Network, DSO, demand/load forecast

1.8. General remarks

General remarks

2. Diagrams of Use Case

3. Technical Details

3.1. Actors

Actor NameActor TypeActor DescriptionFurther information specific to this Use Case
Forecast providerProcessProcess in charge of generating load forecasting.
DSO (Distribution System Operator)System operatorSystem operator in charge of controlling local voltage and power flows in DN.
Coordinet platformIT platformIt receives the load forecast from DSO and shares it (at primary substation level) with the TSO.
SubstationFacilityPrimary or secondary electrical substation where current load data is gathered.
Weather forecasting providerServiceExternal service which receives requests of weather prediction for a moment and returns it.

3.2. References

No.References TypeReferenceStatusImpact on Use CaseOrganistaor / OrganisationLink

4. Step by Step Analysis of Use Case

4.1. Overview of Scenarios

No.Scenario NameScenario DescriptionPrimary ActorTriggering EventPre-ConditionPost-Condition
1Get measurement from substationSee complete descriptionGet measurement from substation
2NoneSee complete descriptionNone
3NoneSee complete descriptionNone
4NoneSee complete descriptionNone


4.2. Steps – Scenarios

Scenario Name:
Get measurement from substation
Step No.Event.Name of Process/ ActivityDescription of Process/ Activity.ServiceInformation Producer (Actor)Information Receiver (Actor)Information ExchangedRequirements, R-ID
Scenario Name:
Step No.Event.Name of Process/ ActivityDescription of Process/ Activity.ServiceInformation Producer (Actor)Information Receiver (Actor)Information ExchangedRequirements, R-ID
Scenario Name:
Step No.Event.Name of Process/ ActivityDescription of Process/ Activity.ServiceInformation Producer (Actor)Information Receiver (Actor)Information ExchangedRequirements, R-ID
Scenario Name:
Step No.Event.Name of Process/ ActivityDescription of Process/ Activity.ServiceInformation Producer (Actor)Information Receiver (Actor)Information ExchangedRequirements, R-ID

5. Information Exchanged

Information exchanged IDName of InformationDescription of Information ExchangedRequirement

6. Requirements (optional)

Category IdentifierNameDescriptionmRID
Req_IDReq_Name‘Load forecast in distribution network’
IE_1Substation measureCurrent load at the substation [KVA]IE_1
IE_2Weather predictionWeather forecasting values: Temperature [oC], relative humidity [%], atmospheric pressure [mbar], cloud index [%], wind speed [m/s], wind direction[degrees], etc.IE_2
IE_3Load predictionLoad Forecast generated by the Forecast Provider [KVA]IE_1, IE_2
IE_4KPIs calculationCalculation of KPIs, done when the moment of the prediction is reached: [KPI_24] and [KPI_25]IE_3

7. Common Terms and Definitions

8. Custom Information (optional)

KeyValueRefers to Section