TREN

Sectors · Tunnel & Underground Construction

SECTOR ENGINE · S-30

Tunnel & Underground Construction

The engine performs one calculation: what the gap between planned and actual TBM advance rate costs in days and in money. The answer carries two numbers and they do not measure the same thing — that difference is most of what this page is about.

54 LINES · 2 METHODS · 1 ROUTE · 1 METHOD NOT CALLABLE

CAPABILITY
  • Advance-rate loss as days and cost
Four numbers go in — planned and actual advance rate, days affected, daily cost — and back come the percentage lost, the extra days and the prolongation cost. To the extent the cause is ground conditions, the contract clause to rest on arrives inside the result too.
THE ENGINE

Where this engine sits

The four things this page rests on: why the sector is one of the four, what the route requires, where the input comes from, and what is not covered here.

Why this sector is one of the four
Tunnelling has an engine named after it, and a route reaches that engine today — that reach is what opened this page. The same reference record also attaches an exactly matching legal branch: tunnels. Two of the four are routed to the nearest branch and carry a note in their record explaining it; here the branch matches one to one, so that note stays blank.
What the calculation requires
The call is tied to a signed-in session: the company identity travels with every request. A call from a browser also carries the session token; a system connecting with its own key is verified by that key. On top of it sits a budget of 60 calls a minute, and that budget belongs to this calculation itself — every calculation keeps its own share.
Who supplies the input
All four numbers come from the caller: planned advance rate, actual advance rate, affected days and daily cost. The engine opens no geotechnical report and connects to no monitoring system. The codebase does hold typical advance ranges by ground type — 15–25 m/day in soft ground, 25–40 in hard rock — but the engine never consults them. Whether a planned rate was reasonable stays outside the engine; what it works on is the gap between the two rates it is given.
What this page covers
The one calculation, how much of its input is actually checked, and the second method nobody can call. The sector’s reference record also holds contract types, tunnel types, construction methods and a geotechnical-risk block. Five of the nine fields are readable from a route: the id and the two names from the sector list, long-lead items and the analysis hooks from the catalog route — the last of them under the name risk drivers. No route reads the other four. None of them is part of this engine.

MEASURED: 54 LINES · 2 METHODS · 1 ROUTE · 9 FIELDS IN THE REFERENCE RECORD, 5 READABLE FROM A ROUTE

All sectors →

THE CALCULATION

Two numbers in one answer

A single call returns both a percentage and a number of days. The percentage measures the rate shortfall; the day count measures the schedule impact. Neither is a restatement of the other, and the gap between them widens as the shortfall grows.

The percentage: rate shortfall

The gap between planned and actual rate, over the planned rate. The divisor is the planned rate.

The days: schedule impact

How many days it adds to drive the same length more slowly. The divisor is the actual rate — that is precisely where it parts company with the percentage.

loss_pct = (planned − actual) / planned × 100

extra_days = (planned − actual) × affected_days / actual

prolongation_cost = extra_days × daily_cost

READINGThe percentage is not a schedule estimate. A 50% rate loss extends the duration by 100%, not 50%. Carrying the percentage into a claim as if it were duration is an error, and the error grows with the shortfall: at a 90% rate loss the schedule impact is ten times the percentage.

MEASURED · ON A 10-DAY WINDOW: 10% RATE LOSS → 11% SCHEDULE · 50% → 100% · 80% → 400% · 90% → 900%

INPUT VALIDATION

What the model rejects, and what it admits

The route’s request model puts a constraint on each of the four fields, and the constraints hold. The one case they do not cover is the one that inverts the calculation.

  • planned advance rate > 0
  • actual advance rate > 0
  • affected days ≥ 1
  • daily cost ≥ 0

RULEThe model does not stop the actual rate from being GREATER than the planned one. On a tunnel advancing faster than planned, the percentage, the extra days and the cost all come back negative. For example, at planned 5, actual 8, 10 days and 1,000 per day the output is −3,750. The inputs are illustrative. The arithmetic is not. The engine has no floor holding the result at zero. The output field’s name ends in _usd, but no currency conversion happens anywhere; the unit is whatever the caller sent.

RULEThe engine’s own guard cannot be reached from the route. The class returns an error dictionary if either rate is zero or negative, but the model rejects those values with 422 first, and that model is the class’s only call site. That error string cannot reach an HTTP caller.

MEASURED: 4 CONSTRAINTS, ALL 4 HOLD · ZERO/NEGATIVE RATE 422 · ACTUAL > PLANNED IS ACCEPTED

THE SECOND METHOD

TBM jamming risk, as a five-question check.

The class’s second method examines who the contract leaves TBM jamming risk with: the ground baseline report’s (GBR) classification coverage, the right to claim under Clause 4.12, and which side bears the recovery cost.

Running on its own, the method returns the five questions below in place of a guess — a checklist in a fixed order:

  • Is TBM jamming defined in the contract?
  • Does the risk sit with the contractor or the employer?
  • Is the GBR soil class comprehensive?
  • Is there a right to claim under Clause 4.12?
  • Who bears the recovery cost?

THE LIST IS A HARD-CODED TURKISH CONSTANT · NO TRANSLATION LAYER TOUCHES IT · SHOWN HERE IN ENGLISH

FIVE FIXED QUESTIONS · THEIR ORDER IS WRITTEN IN THE CODE · THE WHOLE OF THE TBM-JAMMING CHECK

QUESTIONS

Frequently asked

Can the engine tell me how late my tunnel will be?

No. The engine works on the gap between two rates you supply; it does not measure them, predict them, or fetch them from anywhere. You send the planned and actual advance rates, the affected days and the daily cost. The answer is worth exactly what the numbers you sent are worth.

Why don’t the percentage and the day count give the same ratio?

They have different divisors. The percentage divides by the planned rate and gives the rate shortfall; the day count divides by the actual rate and gives how long the same length now takes. A tunnel advancing at half speed has a 50% rate loss but finishes in twice the time. Both are correct, and neither measures the other.

Can I use the TBM jamming risk analysis?

Not today. The method is in the code but there is no route behind it, so nothing outside can call it. We write that here because describing a method that exists without saying it has no route would be an advertisement.

Does the engine read the GBR or the geotechnical report?

No, no document is opened. The sector’s reference record does hold a geotechnical-risk block containing the GBR and GDR definitions, but no route reads that either. All the engine sees is the four numbers in the request body.

Let’s talk about your tunnel claim

If you have advance records and a GBR in hand, we can go through what this engine calculates and what it does not.