Blocking back and queue length at a junction
A capacity ratio tells you how hard an arm is working. Only the queue in metres, compared against measured storage, tells you whether the junction behind will block.
Queue length and blocking back at a junction are the outputs that decide whether a highway authority objects, and they are the outputs most often misread. A capacity ratio tells you how hard a junction arm is working. It does not tell you how long the queue gets, where that queue ends up, or whether it reaches back into the junction behind. Two arms both reporting a ratio of flow to capacity of 0.88 can behave completely differently if one has 90 metres of storage and the other has 25.
So the question to put to your transport consultant is not what the capacity ratio is. It is how long the queue is, in metres, and what it reaches.
What queue outputs actually represent
Standard junction models report queues in passenger car units, and the conversion to metres is a modelling assumption rather than a measurement. A conversion of roughly 6 metres per passenger car unit is common, and the figure used should be stated in the report because a queue of 14 units is 84 metres at that rate and 70 metres at 5 metres per unit.
The queue reported is also a queue at a point in the modelled time period, derived from demand flows averaged over time segments. Models with a single hour peak period smooth out the five-minute surge that actually blocks the junction, which is one reason observed queues on site can be longer than modelled ones even when the model is properly calibrated.
Ask for the time-segment profile rather than the summary table. A junction that reaches its worst queue in the fourth of six segments and recovers is a different proposition from one that is still growing when the model period ends, and the summary row looks identical.
Mean maximum queue and its limits
The mean maximum queue is the average of the maximum queue across the modelled segments. It is a central value, so by definition the queue exceeds it about half the time.
That matters when the number is compared against a storage length. If a right-turn lane holds 35 metres and the mean maximum queue is 32 metres, the lane overflows regularly. Comparing a central estimate against a physical limit, and concluding that the limit is respected, is the single most common error in a Transport Assessment.
Where storage is the constraint, ask for a higher percentile queue or for the distribution across segments. Some authorities will ask for this themselves on a constrained network. Presenting it unprompted is better, because it shows you understood the question.
Blocking back into an upstream junction
Junction models usually treat a junction in isolation. The arm has a demand flow, a geometry and a capacity, and the queue grows upstream into an infinite reservoir. Real roads do not have infinite reservoirs.
Blocking back happens when the queue from a downstream junction reaches the stop line or give-way line of the junction behind, so that traffic in the upstream junction cannot discharge even though that junction has spare capacity of its own. Once it starts, the failure propagates and neither junction's isolated model describes what is happening.
Checking for it is not difficult:
- Measure the storage length between the two stop lines along each movement, from a drawing or a survey rather than from an aerial image.
- Convert each modelled queue to metres using a stated conversion.
- Compare the queue against the storage for every modelled scenario and time segment, not just the worst case.
- Where the queue approaches the storage length, model the junctions as a linked network or in microsimulation.
- Report the comparison as a table of queue against available storage, arm by arm.
The same check applies to a side road queue reaching back into a residential street, and to a queue on an access road blocking a car park or a service yard. Highway authorities worry about these more than about an abstract capacity ratio, because they are what generates complaints.
Capacity ratios and what counts as failure
Conventional practical capacity thresholds are a ratio of flow to capacity of about 0.85 at priority junctions and roundabouts, and a degree of saturation of about 90 per cent at signals. Both are conventions rather than standards, and local highway authorities vary in how they apply them.
Neither threshold is a cliff. What happens above them is that delay and queue respond non-linearly to small changes in flow. Below about 0.85, adding 20 vehicles an hour adds a car or two to the queue. At 0.95, the same 20 vehicles can add ten.
The NPPF test for refusal on highway grounds, in the current edition published on 17 August 2026, remains an unacceptable impact on highway safety or severe residual cumulative impacts on the road network. A capacity ratio above a local convention is not, by itself, a severe residual cumulative impact, and an assessment that explains the difference between exceeding a convention and causing a severe impact is doing useful work.
Sensitivity to small flow changes
Because the response is non-linear near capacity, a model sitting at 0.93 is fragile. A modest change in assumed trip generation, in the distribution of trips across arms, or in background growth moves the queue materially.
Run the sensitivity tests and publish them. The ones worth doing are trip generation at a higher rate than the chosen site selection from the trip rate database, growth to a later assessment year, an alternative turning split at the critical arm, and a committed development added or removed. Four tests, each a row in a table, with the queue and the storage comparison for each.
Where every test shows the queue inside the available storage, the case is strong. Where one test overflows, say so and explain why that test is unlikely or what mitigation covers it. An assessment that reports only the central case invites the authority's consultant to run the sensitivity test for you, and to present the answer less charitably.
Mitigation that actually adds capacity
Options that work tend to involve storage or control rather than width.
Lengthening a flare or a right-turn pocket adds storage directly and is often deliverable within the existing carriageway. Signal timing changes, including re-optimising the cycle and green splits, can move capacity between arms at no capital cost, although the authority owns the timings and will want to see the network effect. Adaptive control on an isolated signal junction handles variable demand better than fixed time.
Banning a turning movement removes the conflict that causes the queue, at the cost of redistributing trips, which has to be modelled rather than assumed. Reducing the trip generation itself through a travel plan is legitimate where the measures are secured and monitored, but authorities discount uncommitted measures heavily.
Where the works fall in the highway, they are delivered under a section 278 agreement, and the way a section 278 agreement follows a permission shapes the programme. On the strategic road network the position is different again, and the grounds on which National Highways objects to an application often turn on queue outputs at the junction with its network.
This is general guidance, not legal advice. Check the position with your local highway authority or a qualified transport planner before relying on it.
Presenting results honestly
A junction assessment that an authority accepts first time usually has six things in it: the survey date with a note on neutrality, the calibration and validation evidence, the growth assumptions with their source, the trip generation method and the site selection used, the queue outputs in both units and metres, and the comparison against measured storage lengths.
An illustrative example
A 180-unit scheme takes access from a priority junction 70 metres upstream of a signalised crossroads. The isolated model reports a ratio of flow to capacity of 0.81 at the access and a mean maximum queue of 4.2 passenger car units at the crossroads approach. Converted at 6 metres, that is 25 metres, comfortably inside 70. The segment profile shows a peak segment queue of 11 units, or 66 metres, which reaches the access. The scheme is revised to relocate the access further upstream. This example is illustrative and not drawn from a real application.
Smaller schemes may need a Transport Statement rather than a full assessment, and the queue question still applies in the same form.
Enter your postcode on PlanScope to see which reports your site is likely to need.
PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. For specialist technical reports, qualified professional input may be required.
Frequently asked questions
What does mean maximum queue actually mean?
It is the average of the maximum queue across the modelled time segments, so it is a central value rather than a worst case. The queue exceeds it roughly half the time. Comparing a mean maximum queue against a physical storage length and concluding the storage is adequate is a common and serious error.
How do you convert a queue in passenger car units to metres?
By applying an assumed length per unit, commonly around 6 metres. The figure is a modelling assumption rather than a measurement, so it should be stated in the report. A queue of 14 units is 84 metres at 6 metres per unit and 70 metres at 5, which can change the conclusion entirely.
Do junction models pick up blocking back automatically?
Usually not. A standard isolated junction model lets the queue grow upstream without limit, as though the road behind had infinite storage. Blocking back has to be checked by comparing the converted queue against measured storage between stop lines, and modelled as a linked network where the margin is tight.
Is a capacity ratio above 0.85 a reason for refusal?
Not by itself. The NPPF test is an unacceptable impact on highway safety or severe residual cumulative impacts on the road network. The 0.85 ratio at priority junctions and 90 per cent degree of saturation at signals are practical conventions, and exceeding one is not the same as causing a severe impact.
Which sensitivity tests should a junction assessment include?
Four cover most ground: a higher trip generation rate than the chosen site selection, growth to a later assessment year, an alternative turning split at the critical arm, and committed development added or removed. Report each with its queue and its comparison against available storage rather than only the central case.
Sources and further reading
This article is general guidance about the planning system in England and, where stated, the rest of the UK. It is not legal or professional advice. Check your local planning authority’s validation requirements and take professional advice for your specific site.