Week 6 Studio: Peak-Load Stress Test

Drive the road network to peak load and watch where it gives way, then argue the fix: a wider road, a better line, or fewer trips needing to be made.

29 March 2027

The 08:00 hold#

For twenty minutes, run the district at in-game 08:00 with the build tools untouched. Describe where the first queue forms, whether it begins at Wickstead or upstream, whether it clears between platoons, and what Hessel Row carries while the corridor loads. Idris Fenn will interrupt anyone reaching for the road tool.

The longest queue is not necessarily the failure point. Walk the Traffic info view north to south and find the stop line that constrains throughput. A still cannot show whether a junction is over capacity or receiving a queue from somewhere else.

Before you touch anything#

Record the 08:00 (peak) and 14:00 (off-peak) route times, volume at Wickstead, daily passengers on the Denton Hill–Barrow Fields spine, and volume on Hessel Row. Pens down at 13:10. The reference route reads 11.2 minutes at 08:00 and 5.6 at 14:00.

Reconstruct the 118% reference capacity claim from the raw hourly count, lane count, supplied saturation-flow assumption and observed green-time fraction. Multiply saturation flow by the green ratio and lanes, then divide the measured demand by that assumed capacity. A percentage without its inputs cannot carry the afternoon’s argument.

Widening Wickstead#

First save the two-lane state as <student ID>-w06-pre. Then widen the Kerrow corridor approaches at Wickstead from two lanes to four. Everybody widens; the continuing course save uses the same prescribed intervention.

Time the route on the first morning, then use the issued kerrow-ref-w06-3w and kerrow-ref-w06-6w checkpoints for the same route and junction count. The reference trajectory is 11.2 → 8.9 → 10.7 minutes: immediate relief, three in-game weeks holding at 8.9, then a partial refill. Do not advance six weeks in one action and conclude that the widening did nothing. It worked, and much of that gain was consumed. The checkpoints preserve that sequence without pretending six weeks fit into forty minutes.

Where the traffic came from#

Audit four refill hypotheses rather than forcing a causal total:

• spatial: volume leaves Hessel Row as the wider route becomes cheaper;
• temporal: the issued 07:00–09:00 count shows trips returning to the peak;
• modal: the transit line’s passenger count changes against last week’s log;
• land use: new occupancy appears at the Denton Hill edge.

For each row, enter an estimate or not estimable, the observation and every conversion assumption. The rows need not sum to an elasticity forecast. If you propose a one-way pair on the corridor and Hessel Row, bring volume, capacity, volume ÷ capacity and frontage land value before and after on both streets.

The argument round#

Four groups take one hypothesis each, put its figure and source view on the board, and re-read any contested row live. Then choose one instrument: converting Wickstead to a roundabout, reallocating a lane to transit, shorter trips through mixed use, or peak pricing. The first three are road-tool moves you can make in the unmodded build; the last one the model has no mechanism for, and signal phasing is not an instrument here either — the build exposes no timing control, and the mod that adds one is outside this studio’s settings. Record both limitations instead of simulating them.

Leave a position statement of no more than 150 words on the wall. Name the instrument, the number that would show it worked, and the trade: a temporary mobility gain against a permanent livability cost on the widened frontage.

Across the break#

Save the widened district as <student ID>-w06, copy it off-machine, and bring it back unrepaired on 19 April. Keep -w06-pre, the three peak figures, the decomposition and your named instrument. The break preparation is on paper: copy the issued utility availability figures, mark the first failure-chronology entry you would investigate, and bring the Week 3 Ombersley zoning memo. No unattended run and no w07-pre file are required.