Why Fault Level Data Breaks Connection Studies
Fault level data is the foundation of every connection study. When the data is wrong, your load flow and short circuit results are wrong too. Most engineers discover data errors after running the study, not before. That costs days of rework.
This guide covers the most common fault level data errors found in DNO exports, a step-by-step validation process, and how to automate the entire check.
Common Errors in Fault Level Data
DNO fault level data arrives in formats ranging from structured CSV exports to hand-edited Excel files. Across UK DNOs, these are the errors that appear most often.
Missing Fault Level MVA Values
The most frequent error is missing or blank fault level MVA fields. This happens when DNOs export partial datasets or when busbars are included as placeholders without calculated values.
- Blank cells in the fault level column where a value between 250 and 2,000 MVA is expected
- Zero values that indicate the field was defaulted rather than calculated
- Placeholder text like "TBC" or "N/A" in numeric fields
Zero or Negative X/R Ratios
The X/R ratio (reactance to resistance) determines fault current decay characteristics. Valid ratios for HV and EHV busbars typically fall between 5 and 15.
| Error Type | What You See | Expected Range |
|---|---|---|
| Zero X/R ratio | 0.00 | 5 to 15 (HV/EHV) |
| Negative X/R ratio | -3.2 | Always positive |
| Extremely high X/R | 150+ | Rarely above 30 |
| Missing entirely | Blank cell | Numeric value |
A zero X/R ratio will cause division errors in short circuit calculations. Negative values indicate data corruption or sign convention errors in the source system.
Inconsistent Busbar Naming
DNO systems use different naming conventions for the same physical busbar. When you merge data from multiple sources, busbar name mismatches create duplicate nodes in your model.
- UKPN might label a busbar as "KINGS_33" while the same busbar appears as "Kings Cross 33kV" in a different dataset
- NGED uses numeric prefixes like "2140_BRIDGWATER_33" while the load flow model uses "Bridgwater 33"
- Voltage suffixes vary: "_33", "_33kV", "33KV", "33 kV"
Stale or Outdated Values
Fault levels change as the network evolves. Data from 2 or more years ago may not reflect recent reinforcement, new generation connections, or decommissioned assets.
- Compare the data date field against the study date
- Check if the dataset version matches the DNO's latest published fault level data
- Flag any values that differ by more than 10% from the previous dataset for the same busbar
Step-by-Step Validation Process
Follow this sequence before importing fault level data into any power system modelling tool.
Step 1: Check File Completeness
Before looking at individual values, confirm the file is complete.
- Count the number of busbars. Compare against the expected number for the DNO region and voltage level.
- Verify all required columns exist: busbar name, voltage level, three-phase fault level (MVA), single-phase fault level (MVA), X/R ratio.
- Check for truncated rows at the end of the file. Some CSV exports cut off at row limits.
Step 2: Validate Ranges
Apply boundary checks to every numeric field.
| Parameter | Minimum | Maximum | Action if Out of Range |
|---|---|---|---|
| Three-phase fault level (MVA) | 50 | 25,000 | Flag for review |
| X/R ratio | 1.0 | 40.0 | Flag for review |
| Voltage (kV) | 0.4 | 400 | Flag for review |
| Single-phase fault level (MVA) | 30 | 20,000 | Flag for review |
Values outside these ranges are not always wrong, but they need a second look. A 400 kV busbar with a fault level below 1,000 MVA, for example, is worth checking.
Step 3: Cross-Reference Busbar Names
Match busbar names against a reference list. If you do not have one, build it from the DNO's published network data.
- Normalize naming: strip spaces, convert to uppercase, remove voltage suffixes
- Use fuzzy matching to catch near-duplicates like "BRIDGWATER" vs "BRIDGEWATER"
- Map each busbar to a unique network node ID
Step 4: Check Consistency Between Fields
Fault level values, voltage, and X/R should be internally consistent.
- Higher voltage busbars generally have higher fault levels
- X/R ratios increase with voltage level (transmission X/R is higher than distribution)
- If three-phase fault level is present but single-phase is missing, flag it
Step 5: Compare Against Previous Data
If you have fault level data from a previous study or dataset version, compare the two.
- Identify busbars where fault level changed by more than 15%
- Check if new busbars appeared or old ones disappeared
- Document changes and verify them against known network modifications
How Noda Automates Fault Level Validation
Noda runs this entire validation process automatically when you upload a DNO fault level file.
- Format detection: Noda identifies the DNO and file format (UKPN, NGED, SSEN, SPEN, NPG, ENW) and maps columns to a standard schema
- Range checks: Every numeric field is checked against voltage-level-specific thresholds, not just global ranges
- Name matching: Noda maintains a reference database of over 50,000 UK busbars and matches incoming names using fuzzy logic
- Version comparison: If you have uploaded a previous dataset, Noda highlights what changed and flags anomalies
- Validation report: A summary report lists every issue found, categorised by severity (error, warning, info)
Instead of spending 2 to 4 hours manually checking a spreadsheet, the validation completes in under 30 seconds.
DNO-Specific Gotchas
Each DNO has quirks in how they export fault level data.
- UKPN: Exports separate files for winter maximum and summer minimum fault levels. Make sure you are using the correct season for your study scenario.
- NGED: Uses internal busbar IDs that do not match published network diagrams. You need a mapping table.
- SSEN: Often provides fault level data as PDF tables rather than structured data. This requires extraction before validation.
- SPEN: Includes both make and break fault levels. Verify which value your modelling tool expects.
- NPG: Data format changed in 2024. Older scripts may fail on the new column layout.
Key Takeaways
- Missing MVA values and zero X/R ratios are the most common fault level data errors across all UK DNOs
- A five-step validation process (completeness, ranges, naming, consistency, comparison) catches 95% of data issues before they affect your study
- Automated validation reduces a 2 to 4 hour manual task to under 30 seconds and eliminates human oversight errors
Next Steps
If your team spends hours checking fault level data before every connection study, there is a faster way. Book a demo to see how Noda validates DNO data files automatically and flags every issue before it reaches your model.
