01
Fast route
Professional Check
Five trace-reading decisions, one integration calculation and a demanding evidence check.
- Five focused lessons
- 10 scored questions
- Answer rationales after submission
About 15 minutes
Vitreous Learning 02
Professional BESS course · Great Britain
Read the trace before you tell the story
Separate intention, instruction, physical outturn and settlement evidence. Then state exactly what the public trace can support.
2 learning routes80% pass standard30 assessment questions
Choose your route
Evidence ledElexonNESOBSCTechnical cut-off · 24 August 2026
One subject · two depths
Choose your route
This course assumes familiarity with the Balancing Mechanism. It concentrates on trace reconstruction, time alignment and evidence limits.
02
Full route · recommended
Detailed Course
From BM Unit identity and Gate Closure through BOALF geometry, settlement revisions and a complete dispatch reconstruction.
- Nine detailed modules
- Greybridge worked case
- 20 scored questions
75 to 90 minutes
Scores and progress remain on this device. No account or sign-up is required.
Professional Check · About 15 minutes
Can you distinguish what was planned, instructed and delivered?
Use a disciplined evidence stack to prevent an attractive chart from becoming an unsupported trading story.
Lesson 1 of 5
Start with identity and time
A trace is unreliable until the BM Unit, settlement date, settlement period and time convention are explicit.
A project name, legal entity and BM Unit identifier may refer to different boundaries. Record the exact BM Unit before joining datasets. Record UTC timestamps as supplied, then map them to settlement periods using the applicable settlement calendar.
Most settlement days contain 48 half-hour periods. Clock-change days can contain 46 or 50. A row number is therefore weaker evidence than the date, period and timestamp together.
Normal day48 periods
Spring clock change46 periods
Autumn clock change50 periods
Hostile check
A clean chart with the wrong BM Unit or a one-hour clock shift can produce a confident, entirely false dispatch narrative.
Lesson 2 of 5
FPN is the baseline
The Final Physical Notification records the expected import or export in the absence of Balancing Mechanism acceptances.
Physical Notifications are submitted as MW points. The prevailing notification at Gate Closure becomes the FPN used in settlement calculations. Integrating the line through the settlement period gives Period FPN in MWh.
The FPN is a notified physical intention. Public data does not reveal every wholesale trade, portfolio transfer or hedge that produced it.
Hostile check
Treating the FPN as the complete trade book converts a physical notification into a claim about private contracts.
Lesson 3 of 5
Read the acceptance geometry
A BOA is an instruction represented by time-stamped MW levels, so volume depends on the area between the accepted level and the baseline.
BOALF records the accepted operating level at the start, end and any intermediate points. A ramp must be integrated. Multiplying the final MW level by thirty minutes usually overstates the accepted volume.
Direction is relative to the notified position. More export or less import is normally an offer direction. Less export or more import is normally a bid direction.
Hostile check
A 40 MW end point reached after a ten-minute ramp does not create 20 MWh of accepted volume across the full half hour.
Lesson 4 of 5
Metered energy is the outturn boundary
Settlement metering reports energy for the period. Divide MWh by period length to obtain average MW.
Eighteen MWh in a half hour equals 36 MW average export. That average can coexist with a 40 MW end level because the unit may have ramped or changed output within the period.
Published metered volumes can be estimated or revised through later settlement runs. Preserve the run type and extraction time in any audit trail.
Metered energy18 MWh
Period length0.5 h
Average export36 MW
Hostile check
A period-average MW value does not reveal the second-by-second trajectory or prove compliance with every instruction point.
Lesson 5 of 5
Bound the state-of-charge inference
A public dispatch trace supports an energy-flow estimate when the initial state, efficiency boundary and auxiliary load are declared.
Integrating metered import and export can show net grid-boundary energy. State of charge requires additional assumptions about conversion losses, auxiliary demand, operational reserves, internal curtailment and the starting condition.
Label the conclusion as observed, calculated, inferred or unknown. This prevents an energy-flow calculation from being presented as direct battery telemetry.
Hostile check
A plausible state-of-charge curve is still an inference unless the starting state and internal measurement boundary are known.
Primary and standards sources used in this course
Professional Check assessment
Answer every question before submitting. Correct answers and rationales appear only after the complete attempt.
Professional CheckQuestion 1 of 10
Detailed Course · 75 to 90 minutes
Reconstruct a dispatch trace from first principles
Follow one BM Unit from identifiers and timestamps to FPN, accepted levels, metered energy and a bounded conclusion.
Module 1 of 9
Build the evidence stack
Keep identity, intention, instruction, outturn and settlement in separate layers until the joins are proven.
Start with BM Unit registration and the project-to-unit mapping. Add FPN, Bid-Offer Data, BOALF and metered volumes as separate tables. Preserve source, extraction time, units, sign convention and settlement run.
Only then derive period averages, accepted volumes and bounded state-of-charge movements. Every calculated field should point back to the exact source rows used.
Hostile check
Combining feeds first and documenting them later hides duplicate rows, mismatched units and corrections.
Module 2 of 9
Control the clock
GB settlement data combines settlement dates, period numbers and UTC timestamps, with clock-change days requiring explicit treatment.
Use RFC 3339 timestamps where the source supplies them. Convert for display only after calculations. Test that every acceptance lies within the intended settlement date and period, including actions that cross a period boundary.
Keep delivery time, acceptance time and publication time separate. They answer different questions about instruction, action and data availability.
Hostile check
A local-time chart can silently duplicate the autumn hour or omit the spring hour.
Module 3 of 9
Reconstruct the FPN line
Point FPN values are MW spot points and the FPN between them is obtained by interpolation.
Integrate the interpolated line to calculate the Period FPN. A flat 20 MW export across thirty minutes implies 10 MWh. A linear ramp from 0 to 20 MW across the period implies 5 MWh.
Check revisions and data status. A late correction can change the baseline against which accepted volumes are settled.
Flat 20 MW10 MWh
0 to 20 MW ramp5 MWh
BoundaryGrid MW
Hostile check
Using a simple average is valid for a linear segment. It is unsafe when intermediate points change the shape.
Module 4 of 9
Separate prices from instructions
Bid-Offer Data contains price and level information. A BOA records which action NESO accepted.
Pair the acceptance with the correct bid-offer pair and effective data. Check acceptance number, BM Unit, times and flags. A price row without an acceptance is an available action, not a dispatched action.
Selection can reflect location, ramping and system need. Price order alone cannot reconstruct NESO's operational decision.
Hostile check
Matching on BM Unit and settlement period alone can attach the wrong price when multiple pairs or revisions exist.
Module 5 of 9
Integrate BOALF correctly
Accepted volume is the signed area between accepted operating level and the FPN over time.
Split the geometry wherever either line changes slope. Integrate each linear segment using the trapezoid rule and sum the results. Preserve sign convention before classifying bid or offer volume.
For a ten-minute linear rise from 0 to 30 MW, followed by twenty minutes at 30 MW, energy above a zero FPN is 2.5 MWh plus 10 MWh, giving 12.5 MWh.
Ramp energy2.5 MWh
Flat energy10 MWh
Accepted volume12.5 MWh
Hostile check
Thirty megawatts multiplied by half an hour gives 15 MWh and overstates this ramped action by 2.5 MWh.
Module 6 of 9
Use metered volumes with settlement lineage
Metered data is period energy and later settlement runs can replace earlier estimated values.
Record whether the value is actual or estimated and which run supplied it. A public analysis should state its as-of date because II, SF and reconciliation data can differ.
Compare metered energy with Period FPN and accepted volume as a reconciliation, while allowing for imbalance, other instructions, losses and data corrections within the defined boundary.
Hostile check
A difference between metered volume and FPN plus BOA is a reconciliation item. It is not automatically evidence of non-delivery.
Module 7 of 9
Infer state of charge conservatively
Grid-boundary power can estimate energy movement, while internal state requires declared efficiency and auxiliary-load assumptions.
Apply charge and discharge efficiency on the correct side of the boundary. Include auxiliary consumption where material. Carry an interval or range when the starting state is unknown.
Respect operational reserves and energy windows. The physical cell capacity can exceed the dispatchable window exposed by the control system.
Hostile check
A calculated cumulative MWh curve can drift materially from true state of charge over many cycles when losses and auxiliaries are ignored.
Module 8 of 9
Reconcile cashflow without inventing profit
Accepted volume and price can support a BM cashflow component. Profit requires the underlying energy position and costs.
Calculate each signed acceptance cashflow using the applicable settlement treatment. Keep the associated wholesale position, imbalance and rebalancing outside the public result unless supported.
Use a bridge with observed BM receipts or payments, calculated volumes, inferred energy effects and unknown private elements. This makes the boundary visible.
Hostile check
Calling a positive offer receipt profit ignores the acquisition value of stored energy and every private offset.
Module 9 of 9
Greybridge worked reconstruction
A defensible trace ends with a bounded conclusion and a list of unresolved evidence.
Greybridge has a zero FPN. Its BOALF rises linearly from 0 to 40 MW over five minutes, holds 40 MW for twenty minutes, then falls to zero over five minutes. The accepted energy is 1.67 + 13.33 + 1.67 = 16.67 MWh.
Settlement metering later reports 18 MWh export. The trace supports a 36 MW period average and a 1.33 MWh reconciliation difference. Possible causes remain to be tested against additional instructions, baseline corrections, metering lineage and imbalance.
Accepted energy16.67 MWh
Metered export18 MWh
Reconciliation1.33 MWh
Hostile check
The 1.33 MWh difference must remain unresolved until evidence assigns it. Choosing the most convenient explanation is not analysis.
Primary and standards sources used in this course
Detailed Course assessment
Answer every question before submitting. Correct answers and rationales appear only after the complete attempt.
Detailed CourseQuestion 1 of 20