Customers / Case study
A permit model that stands up at inquiry
The consultancy prepares the air quality assessments that industrial sites need for a permit. When an objector’s expert questioned a plume, the firm could show every input and rerun the model in front of the inspector.
The result in three figures
-
214
Permit scenarios rerun overnight when the meteorological data changed
-
−60%
Time spent on model quality assurance
-
1
Public inquiry at which the model was run live
The situation
An air quality assessment for a permit turns an emissions inventory, a stack layout and five years of meteorological data into a map of concentrations. The regulator reads the map. An objector, if there is one, reads the inputs, and a small error in one of them, such as a stack height, a building outline or a wind sector, can change what the map shows.
The consultancy’s model files had grown by project and by consultant. Checking one meant opening it, and checking that the checker had opened the right one meant a second person.
What changed
GraphAtlas AERMOD reads the inputs as data and runs the dispersion procedure on the kernel, so that the same inputs give the same map on any machine. Each run ends in a receipt that names the inputs, and the web interface lets the user change one of them and see the effect at once.
When the meteorological data for a region was reissued, the firm reran 214 permit scenarios overnight and compared the receipts of the old and the new run, which showed it which of the 214 had moved and by how much.
The result
Quality assurance of the models takes 60% less time, because it is a comparison of receipts and a review of what differs. At a public inquiry in the spring the objector’s expert asked for the height of a stack to be changed. The firm changed it and ran the model in the room, and the inspector took the result into the record.
At the inquiry, the other side’s expert asked us to change a stack height. We changed it and ran the model in the room. Nobody there had seen that done.
What they run
The engines behind it
This customer runs one engine from GraphAtlas, on the same kernel. Each page shows the calculation, the clauses it covers and the evidence for it.
Keep reading
More from our customers
- Mid-size EU asset manager PRIIPs 1,946 Share-class documents refreshed each quarter 1,946 key information documents, with the arithmetic shown Read the study
- US cryptographic module vendor Entropy validation, PQC 0 Digits of difference from the laboratory’s estimator output Agreeing with the laboratory to the last digit Read the study
- Major UK bank SIMM, FRTB-SA 5.8 h Median time to settle a margin dispute, from 4.2 days A margin dispute that took four days now takes an afternoon Read the study