At a distribution company, an outage record often starts as a single line: which feeder tripped, when it opened and when it was re-energized. In supply continuity reporting, though, the real questions sit behind that line. When the feeder opened, which installations lost supply? How long did each of them wait? And how much energy failed to reach customers during that time?
The answers feed the tables submitted to EPDK, the continuity indicators and the compensation owed to consumers. If they are wrong, every report built on them is wrong too. This article looks at why affected installations and energy not supplied should not be counted by hand, and what changes when they are calculated from the network topology.
What Does Supply Continuity Reporting Measure?
Turkey's regulation on supply continuity and commercial and technical quality in distribution (Elektrik Piyasasında Dağıtım Sistemi Tedarik Sürekliliği, Ticari ve Teknik Kalite Yönetmeliği) requires distribution companies to record, classify and regularly report interruptions. Outages are split into planned and unplanned; for planned outages, consumers are expected to be notified in advance. The aim is to make visible both network-wide performance and the experience of each individual consumer.
Internationally, this performance is usually expressed through indicators such as SAIDI (average interruption duration per customer) and SAIFI (average interruption frequency per customer). Both share the same numerator ingredients: the number of customers affected and how long they were without supply. Affected-installation data is therefore the most basic input to the whole reporting chain.
The Hidden Risks of Counting by Hand
Many companies still derive the number of affected installations by estimation or by working through spreadsheets. A typical approach is to pull the customer list per transformer, mark the transformers involved and add them up. It looks practical, but it carries several serious risks:
- Stale connection data: Newly connected, relocated or transferred installations stay on the list in their old position.
- Missed switching: Load transfers during an outage mean some installations experience a shorter interruption, or none at all. A manual count usually loses that distinction.
- Staged restoration: After a fault, the network is typically restored section by section. Writing the same duration against every installation can make indicators look worse or better than reality.
- Results you cannot reproduce: Two people counting the same outage may get two different numbers, and in an audit there is no clear answer to "how was this figure reached?"
Each of these errors can look minor on its own. Across thousands of outage records in a year, they directly undermine the reliability of the report and the accuracy of compensation.
What Network Topology Changes
Topology describes how network elements are connected: which breaker feeds which feeder, which line runs to which transformer, which installation hangs off which network point. Once an outage is recorded as one or more elements opening on that structure, affected installations are no longer counted; they are derived from the connectivity.
This is where staged outage records matter. When each stage records which network element was opened and which switching operation was performed, the period each installation spent without supply emerges installation by installation. Interruption duration is then determined per installation rather than once for the outage as a whole.
The biggest gain is consistency. For the same outage, the same network state and the same stages, the result is the same every time. And the path to the result is traceable: which element opened, which installations sat below it, which switching operation restored which of them.
Energy Not Supplied and Load Profiles
Energy not supplied is the amount of energy that could not reach consumers during an interruption. The simplest estimate multiplies an installation's average consumption by the outage duration. But consumption is not flat across the day. A household draws differently in the evening than after midnight, and an industrial site behaves differently during working hours than at the weekend.
Calculating energy not supplied with the consumer group's load profile therefore gives a more realistic figure. Two outages of the same length, at different times of day, correspond to different amounts of undelivered energy. A profile-based calculation captures that difference and gives more meaningful data for investment prioritization and for assessing network performance.
Getting Master Data Right: Installations and Network Connections
Calculation from topology is only as reliable as the way installations are connected to the network. Installation master data and installation-to-network connections are therefore an integral part of any supply continuity effort. Being able to bulk-load these connections from a file makes both the initial setup and periodic updates far easier.
Personal data is another consideration. Installation records hold consumer information. Storing it encrypted, showing it masked on screen and logging every access is how a company meets reporting needs and data protection obligations at the same time.
From Report to Map: Seeing the Result
The calculated impact should not stay in tables. Showing outages, affected installations and network elements on the same map lets teams sanity-check the result visually. Drawing an area to filter outages in a region shows where recurring problems concentrate and lays the groundwork for identifying the worst-performing network elements.
Conclusion: Calculate Affected Installations from Topology with GeoEner Tedarik Sürekliliği
GeoEner Tedarik Sürekliliği (GeoEner Supply Continuity) records every outage with its stages, notification, status workflow and change history, keeping the network element opened and the switching performed at each stage. Affected installations are not counted by hand; they are calculated from the GeoEner network topology. Interruption duration and energy not supplied are computed per installation using the consumer group's load profile, and installations and their network connections can be bulk-loaded from a file.
From this data, EPDK Tables 1, 3, 4, 5 and the Customer Interruption Record are produced, cross-table consistency is checked and results export to Excel. The outage map shows the impact on the network, and installation personal data is kept encrypted and masked. If you would like to see how affected-installation calculation works on your own network data, get in touch and we will set up a demo.
