There is a particular kind of loss that only happens on pavement work. The job goes well. The material arrives, the layers go down, the levels are right, the surface looks correct. Then the results come back and a lot fails, and the argument about what happens next runs for a fortnight while the surfacing crew you booked sits idle and the following layer cannot proceed.

Pavement construction is the part of civil work where the acceptance criteria are most explicit, most measurable and least negotiable. Earthworks tolerances have some give. Drainage either drains or it does not. A pavement is accepted against numbers produced by a laboratory, and either the numbers are inside the limits or the layer comes out.

That makes it unusually unforgiving of a contractor who prices the placing and not the passing. This guide covers the technical and commercial spine of getting a pavement accepted: subgrade preparation, the pavement stabilisation options and their risks, the testing and lot regime, and the conformance record that turns completed work into a paid claim. It is a delivery guide — for how to construct the pavement section of a tender response, see our guide to winning tenders for road construction and resurfacing.

The pavement as a system, not a set of layers

A road pavement distributes wheel loads down through progressively weaker, cheaper materials until the stress reaching the natural ground is low enough that it does not deform. Every layer depends on the one beneath it.

LayerWhat it doesWhere contractors lose money
SubgradeThe natural or prepared ground the pavement sits onAccepting a soft subgrade under programme pressure and paying for it in every layer above
Selected subgrade / cappingImproves a weak subgrade to a workable platformNot identified at tender because it only appears in a note on the drawings
SubbaseStructural layer, often lower-specification materialMaterial substitution assumed at tender and not approved in delivery
BaseThe main structural layer, highest specificationMoisture and compaction failures; the layer where most retesting happens
SurfacingWaterproofs and provides the running surfaceBooked to a date the base cannot support; the cost of a delayed seal

The systemic point has a direct commercial consequence. A problem in a lower layer does not stay in that layer. A subgrade that will not compact produces a subbase that will not compact, and no amount of effort on the base will fix a foundation problem underneath it. Contractors who lose money on pavements usually lost it several layers before the failing test.

The corollary is worth stating plainly, because it runs against the instinct to keep moving: the cheapest moment to fix a pavement problem is always the earliest one. Every layer placed over a defect multiplies the cost of correcting it.

Who owns the design, and what you may change

On most civil work the pavement is designed by the client’s engineer, and you build what is drawn. That does not make the design irrelevant to you — it makes it something you have to read for risk rather than for method.

  • What subgrade strength does the design assume? It will be stated as a design value. If the actual subgrade does not achieve it, the pavement above is under-designed and someone has to decide what to do. Whether that is your problem depends on the contract — see latent conditions in civil contracts.
  • Is a pavement investigation report attached, and how many test locations does it contain? A design based on three test pits over two kilometres is a design based on assumption, and the assumption is being transferred to you.
  • Are the layer materials specified by product or by property? A specification calling for a named road authority material class is a supply constraint; one calling for properties may allow a local alternative and is worth investigating with your quarry.
  • Is stabilisation designed, or is it a contingency? The difference is substantial. Designed stabilisation is priced work. Contingent stabilisation — “if the subgrade is unsuitable, stabilise” — is an open-ended obligation with no quantity.

Where you carry the design, the position reverses and the exposure is larger. Under a design and construct arrangement you own the design assumptions, including the subgrade assessment, and a pavement that fails in service is a defects and potentially a fitness for purpose question — the risk transfer explained in our guide to contract forms beyond construct-only. Contractors who take design responsibility for a pavement without an adequate geotechnical investigation have taken a risk they cannot quantify.

Subgrade: the layer that decides everything above it

Subgrade preparation is where pavement jobs are lost, and it is the layer most likely to be treated as part of earthworks and therefore priced by someone who was not thinking about the pavement.

The condition you need is not complicated to describe: a subgrade at an acceptable moisture condition, compacted to specification, at the right level, that does not deform under load. The difficulty is that several common civil situations make it hard to achieve.

  • Wet reactive clay. Above optimum moisture it will not compact and cannot easily be dried in the field. It is the single most common cause of pavement programme collapse in southern and eastern Australia.
  • Cut-fill transitions, where the material changes across a short distance and behaviour changes with it.
  • Existing pavement in a rehabilitation job, where what is under the seal is genuinely unknown until it is opened — the situation in most council road renewal work.
  • Services in the subgrade. Shallow services restrict compaction plant and produce localised soft spots that will not be resolved by more rolling.
  • Trench reinstatement across the alignment, which is a different material at a different age and will behave differently — the interface described in our guide to utility and telecommunications civil works.

The commercially important question is what you are entitled to do when the subgrade is not acceptable, and the answer comes from the contract rather than from engineering. Undercut and replace, stabilise, or re-design are three different costs with three different programme impacts, and which of them is a variation depends entirely on whether the condition was foreseeable and what the documents said about it. Establish that before you start, not when the surveyor tells you the level has dropped 300 millimetres.

Proof rolling, and what a failure costs

A proof roll is a loaded vehicle driven over the prepared surface while the superintendent’s representative watches for deflection or movement. It is a visual, judgement-based acceptance test, and it usually sits on a hold point — meaning no further work proceeds until it passes.

Three things about proof rolls that matter more commercially than technically.

  • It is a judgement call, which makes attendance and agreement essential. Have the right person present, agree what is being assessed and where, and record the outcome and the location in writing on the day. A disputed proof roll a week later is unresolvable.
  • A failure is expensive out of all proportion to the area involved. A soft area over a short length can stop the entire following operation, and the plant standing while it is treated is being paid for.
  • It has a weather dependency. A surface that passed on Thursday may fail on Monday after rain. Passing a proof roll and then leaving the subgrade exposed over a wet weekend is a common and entirely avoidable loss.

The practical control is to sequence so that a passed subgrade is covered quickly, ideally the same day. That has consequences for how you plan the works and how much subgrade you open at once, and it is the sort of method logic that belongs in the programme rather than being discovered on site — see writing a construction programme for tenders.

Stabilisation: the options and when each is used

Soil and pavement stabilisation modifies material in place so that it performs better than it naturally would. It is used either to make an unacceptable subgrade acceptable, or to build a structural layer from material that would otherwise be unsuitable — including the existing pavement, which is why it is central to road rehabilitation.

TypeWhat it doesTypical application
LimeReacts with clay to reduce plasticity and moisture sensitivity, improving workability and strengthWet or reactive clay subgrades — the classic subgrade improvement treatment
CementBinds granular material into a bound layer with significant strength gainSubbase and base construction, and upgrading marginal granular material
Lime-cement and other blendsCombines plasticity modification with strength gainMaterials with both clay content and a structural requirement
Foamed or emulsified bitumenProduces a flexible bound layer, often with a small amount of added lime or cementIn-situ pavement recycling on rehabilitation projects
Granular / mechanicalBlending in a different material to correct gradingWhere the deficiency is grading rather than plasticity or strength

Two distinctions do most of the work in understanding these.

  • Modification versus bound stabilisation. Modification changes the material’s behaviour — it becomes workable, less moisture sensitive, easier to compact — without creating a rigid layer. Bound stabilisation creates a layer with real tensile strength, which changes how the pavement behaves structurally and introduces shrinkage cracking as a consideration. They are different products with different design intents, and the specification will say which is required.
  • In-situ versus plant-mixed. In-situ stabilisation is performed by a purpose-built machine working the existing material in place. Plant-mixed material is produced elsewhere and delivered. In-situ is cheaper and faster where it suits, and is the standard approach for pavement recycling; it also means the material you are working with is whatever is actually there.

Stabilisation is usually performed by specialist subcontractors with dedicated plant, and this is one of the few civil operations where the specialist market is genuinely thin in some regions. Availability of a stabilisation crew is a programme risk in its own right, and one worth confirming before you commit to a date rather than assuming a machine can be found.

The stabilisation risks nobody prices

Stabilisation looks like a straightforward unit-rate operation and behaves like a chemical process with a deadline. Five risks recur.

  • Working time. Once binder is added, there is a limited window to mix, level and compact before the material begins to set. Run past it and the layer does not achieve density and has to be reworked or removed. This drives everything about how the day is planned, and it is why a stabilisation run that starts late is often better postponed than pushed.
  • The design mix is not the field mix. Binder content comes from laboratory testing of samples. The material in the ground varies. Where it varies significantly along the alignment, a single binder rate produces over-treated and under-treated sections, and the under-treated ones fail.
  • Weather. Rain during or shortly after a run can ruin a layer that has not yet gained strength. Wind affects spreading and dust control. The weather dependency is severe and short-notice.
  • Curing and trafficking. A bound layer needs to gain strength before it is loaded. Site traffic driving over a fresh layer to reach the next work front is a common cause of damage, and it is entirely self-inflicted.
  • Depth control. The machine mixes to a set depth. Too shallow leaves untreated material in the layer; too deep dilutes the binder with material that was not meant to be in the mix. Verification during the run, not after, is the control.

There are two further considerations that sit outside the technical envelope but reliably generate complaints and, occasionally, stop-work directions. Lime and cement spreading generates dust, which on a site near houses, a school or a hospital is a community issue and sometimes a regulatory one — see noise, vibration and dust management. And the binders are hazardous materials, with genuine health risks from inhalation and skin contact that require real controls rather than a line in the SWMS.

Moisture is the whole job

If there is one thing that determines whether a pavement layer passes, it is moisture content at the time of compaction. Materials compact properly within a range around optimum moisture. Outside it, no amount of rolling achieves density.

  • Too dry is generally fixable — add water, mix, allow it to distribute. It costs time and water cart hours.
  • Too wet is the expensive direction. Drying material in the field depends on weather you do not control, and on a clay it can take days. This is why wet-weather delay on a pavement job is not just the days it rained.

Three operational implications follow, and they are the difference between a crew that consistently passes and one that does not.

  • Test moisture before compacting, not after failing. A field moisture check takes minutes. Retesting a failed lot, and reworking it, takes a day.
  • Protect the surface. Sealing the surface with the roller at the end of each day so water runs off rather than in, maintaining crossfall on the working platform, and not leaving a layer open over a forecast rain event.
  • Do not open more than you can close. The single most reliable predictor of moisture trouble is the amount of pavement exposed when the weather turns.

None of this is unknown to an experienced pavement crew. It is included here because it is almost never reflected in the tender programme, which typically shows layers progressing at a constant rate with a weather allowance expressed as a number of days. That is not how pavement work behaves, and a programme that does not reflect it is a programme that will be behind.

Lots, sublots and why the definition matters commercially

Pavement work is accepted in lots. A lot is a defined quantity of work — a length, an area or a volume of one material, placed in one layer, under uniform conditions — that is tested and accepted or rejected as a unit. Lots are often divided into sublots for the purpose of locating tests.

Contractors treat lot definition as a quality function. It is a commercial decision, and it is one of the few genuine levers you have on pavement risk.

 Large lotsSmall lots
Testing costLower — fewer tests per unit of workHigher
AdministrationLess paperworkMore records, more lot sheets
Consequence of a failureThe whole lot is in question — potentially a large area of reworkContained; the failure affects a small area
Best used whereConditions are uniform and the crew’s results are consistentMaterial or conditions are variable, or early in a job before results are established

The sensible practice is to start with smaller lots until the process is demonstrably in control, then increase lot size — and to break a lot at any point where conditions genuinely change: a different material source, a cut-fill transition, a change in weather, a different day, a section over a services trench. A lot that spans a known change is a lot that has a failure inside it waiting to be found.

One trap worth naming. Do not include an area you have doubts about inside a larger lot in the hope that the averaging carries it. Where the specification includes a characteristic value or averaging provision this can look attractive, and it occasionally works. When it does not, you have converted a small rework into a large one, and you have a conformance record that shows you knew.

The testing regime and who pays for it

The specification sets out what is tested, how often, by what method, and to what acceptance criteria. The tests generally cover material properties, compaction and moisture, layer thickness and level, and in some cases the strength or stiffness of the completed layer.

Rather than reproducing frequencies or limits — which are specification-specific and vary between road authorities and contracts — the commercially significant questions are these.

  • Who engages the laboratory, and is it required to be accredited? Most road authority specifications require testing by a laboratory with accreditation for the specific test methods. Using a laboratory without the right accreditation produces results that are not acceptable, and the work has to be retested.
  • What is the turnaround? This is the question that decides your programme. Some tests are same-day; others take days. Laboratory turnaround is a programme constraint, and it is one of the most under-recognised on pavement jobs — it is also the reason regional work can be slower, because samples travel.
  • Who pays for retesting after a failure? Almost always you, and it is usually stated. Retesting cost is a real allowance, not an exception.
  • Is there a nominated or client-appointed laboratory? If so, you are dependent on a resource you do not control, and its availability is a risk to log — see the tender risk register.
  • What testing do you do for your own purposes? Process control testing — checking moisture and density before calling for acceptance testing — is not usually specified and is the cheapest insurance available on a pavement job.

That last point deserves emphasis. The contractors who rarely fail a lot are the ones testing themselves before the acceptance test. A nuclear density gauge on site, used by the crew as a process tool, converts a pass-or-rework gamble into a controlled operation. The cost is trivial against a single failed lot.

Hold points, witness points and the programme

Hold points and witness points structure the inspection regime, and they sit in the inspection and test plan covered in our guide to quality management plans and ITPs. On pavement work they have unusual programme force because they cluster at layer boundaries — exactly the moments when the following operation is waiting.

  • A hold point stops work until released. Proof roll acceptance, subgrade approval before placing, and layer conformance before the layer above are common pavement hold points.
  • A witness point allows work to proceed if the nominated party does not attend after proper notice — which makes giving proper notice, in the required form and timeframe, the whole mechanism.

Three practical points that recover real time.

  • Give notice in writing, in the specified form, with the specified lead time, every time. Verbal notice to the person standing next to you is how a witness point becomes a dispute.
  • Log delays in release. Where the superintendent’s representative does not attend or does not release within a reasonable time, that is a delay event and it needs to be recorded contemporaneously to be claimable — the discipline in our guide to extension of time and delay claims.
  • Map the hold points onto the programme before you start. A pavement programme that does not show test turnaround and release time between layers is showing a sequence that cannot happen.

Non-conformances and the four ways they end

A failed test produces a non-conformance, and there are only four outcomes. Knowing them in advance changes how you respond on the day.

OutcomeWhat it meansCost profile
ReworkRecompact, remix, add moisture, re-treat and retestUsually the cheapest if caught immediately, and expensive once covered
Remove and replaceThe layer comes out and is rebuiltThe most expensive, including disposal and replacement material
Accept with a concessionThe client accepts non-conforming work, sometimes with a reduced payment or extended warrantyNegotiated, and entirely at the client’s discretion
RedesignThe pavement above is adjusted to accommodate the layer as builtRequires the designer, takes time, and is uncommon on small work

Four things to get right when one occurs.

  • Establish the extent before agreeing anything. One failing test does not necessarily condemn a whole lot; additional testing may establish that the problem is localised. That investigation is usually worth its cost.
  • Find the cause, not just the fix. A moisture failure caused by a leaking water main under the alignment will recur, and it may not be your risk.
  • Do not concede cause to secure a quick resolution. Where the failure may stem from a design assumption, unforeseen ground conditions or a client-supplied material, the entitlement question and the fix question are separate and should be handled separately.
  • Close it properly. A non-conformance that is fixed but not formally closed leaves a gap in the conformance record, and that gap will be found at the end of the job when you want your money.

The surfacing interface

The handover from pavement to surfacing is a contractual and practical seam, and it produces a predictable set of problems.

  • The surfacing contractor will not accept a base that does not meet their requirements — shape, texture, cleanliness, moisture and level. If they do accept it and the surface fails, the argument about cause is unpleasant and expensive.
  • Sprayed seal work has weather and temperature constraints that are narrower than most contractors allow for, and in southern states there is effectively a sealing season. Missing it can mean leaving a pavement unsealed over winter, which is its own risk and cost.
  • Asphalt has its own temperature and delivery constraints, and a plant that supplies to a schedule.
  • Surfacing crews are booked in advance and charge for standing. A pavement that is a day late is a surfacing crew that is a day idle, and that cost usually lands on you.
  • An unsealed pavement deteriorates. Exposed base loses fines, takes up water and is damaged by traffic. Time between conformance and surfacing is not free.

The practical consequence is that the surfacing date works backwards through your whole programme, and pavement conformance has to be achieved with enough margin that a single failed lot does not put the seal date at risk. Contractors who plan to achieve conformance the day before the sealing crew arrives are planning to pay for a standing crew. Our guide to sprayed sealing and bituminous surfacing covers what that crew is waiting to do.

The conformance pack, and why it is a payment document

The conformance record is the compiled evidence that the pavement was built as specified — lot registers, test results, survey conformance, material certificates, hold point releases, non-conformance reports and their closeouts.

Contractors treat this as an end-of-job administrative task. It is a payment document. Progress claims, practical completion, release of retention and the start of the defects period all depend on it, and the connection is direct: incomplete conformance records are one of the most common reasons a final claim stalls, as covered in our guide to practical completion, defects liability and the final claim.

  • Compile as you go, lot by lot. Reconstructing which test belongs to which lot six months later is genuinely difficult and sometimes impossible.
  • Keep the lot register live — location, material, date, tests called, results, status. It is the index to everything else.
  • Chase results. A laboratory result that never arrived is a hole in the pack, and nobody notices until the end.
  • Close every non-conformance formally, with evidence of the corrective action and the retest.
  • Match survey conformance to the lots, so levels and thickness are evidenced against the same areas as the density results.

There is a second-order benefit. The conformance pack is the most credible past-performance evidence a civil contractor can produce, and very few produce it. A referee statement says you did the job; a clean conformance record shows how you ran it — material worth capturing systematically for the purposes described in our guide to referees and past project experience.

Pricing pavement work honestly

The items that get missed are consistent, and most of them relate to passing rather than placing.

  • Testing. Both the specified acceptance testing where you bear the cost, and your own process control testing. Include retesting at a realistic rate rather than zero.
  • Water. Water cart hours, and the cost and availability of the water itself. On some sites water supply is a genuine constraint and a licensing question — see dewatering, water take and discharge.
  • Moisture conditioning time. Not just placing rates but the time to bring material to condition, which is weather-dependent.
  • Layer protection and re-preparation. Trimming and re-sealing a surface that has been open, after weather or trafficking.
  • Stabilisation mobilisation. Specialist plant charges for mobilisation and minimum quantities; a small stabilisation run can cost disproportionately.
  • Binder. Supply, delivery, storage, and the wastage from spreading.
  • Standing time at hold points, which is real even where it is not claimable.
  • Survey. Set-out and conformance survey are separate activities and conformance survey is frequently forgotten.
  • Traffic management, which on pavement work in a live road is a major cost and often the largest single preliminary — see traffic management plans.
  • The rework allowance. On a pavement job, some rework is close to certain. Pricing zero rework is pricing a perfect job.

The general estimating discipline in our guide to preparing civil works cost estimates applies. The pavement-specific point is that the production rate is not the constraint — the acceptance cycle is, and an estimate built purely from placing rates will be short.

What a pavement tender response has to show

Evaluators on pavement work are usually technical, and they are reading for whether you understand the acceptance regime. Five things distinguish a strong response.

  • A sequence that reflects the hold points, with test turnaround and release shown rather than assumed away.
  • A stated position on subgrade risk — what you have assumed, what you will do if it is worse, and how that is identified early rather than at proof roll.
  • A lot strategy, with the reasoning. Very few bidders offer this and it demonstrates real quality thinking rather than a generic quality statement.
  • Process control described as distinct from acceptance testing. This is the clearest available signal that a contractor manages pavement work rather than hoping.
  • Weather and moisture management as a method, not an allowance — how much is opened at once, how surfaces are protected, what the wet-weather response is.

And the failure modes, which the methodology guidance in our guide to writing a construction methodology statement covers generally: describing the layers rather than the process, quoting the specification back at the evaluator, and offering a programme in which every layer proceeds immediately after the one below with no testing cycle in between. That last one is the tell, and technical evaluators recognise it instantly.

Checklist

  • Do you know the design subgrade strength assumption and how it will be verified?
  • How many geotechnical test locations sit behind the pavement design, and over what length?
  • Is stabilisation designed and quantified, or contingent and open-ended?
  • Are layer materials specified by product or by property, and have you checked local supply against both?
  • Do you know what you are entitled to do, and be paid for, if the subgrade is unacceptable?
  • Is the proof roll attendance, agreement and record process settled before the first one?
  • Is the sequence planned so a passed subgrade is covered quickly?
  • Have you confirmed availability of a stabilisation crew for your dates?
  • Is the working time after binder addition reflected in how each run is planned?
  • Is binder depth verified during the run rather than after?
  • Is dust from binder spreading addressed for the neighbours as well as the workforce?
  • Are you testing moisture before compacting rather than after failing?
  • Is the amount of pavement open at any time limited by what you can close?
  • Have you set a lot strategy, starting small and breaking lots at genuine changes in condition?
  • Is the laboratory accredited for the specific test methods, and do you know the turnaround?
  • Is laboratory turnaround shown in the programme between layers?
  • Are you doing your own process control testing separate from acceptance testing?
  • Is hold point notice given in writing, in the specified form and lead time, every time?
  • Are delays in hold point release logged contemporaneously?
  • Is the conformance pack compiled lot by lot as the work proceeds?
  • Is every non-conformance formally closed with evidence of corrective action and retest?
  • Does the programme leave margin between pavement conformance and the booked surfacing date?
  • Have testing, retesting, water, moisture conditioning, conformance survey and a realistic rework allowance been priced?

Sources and further reading

This guide is general information for Australian civil construction businesses and is not engineering, geotechnical, materials or contractual advice. It deliberately states no compaction limits, moisture ranges, binder contents, layer thicknesses, testing frequencies, curing periods or acceptance criteria; those values are set by the pavement design, the project specification and the relevant road authority standards, they differ between jurisdictions and contracts, and they must be taken from the documents for your project. Pavement design, stabilisation binder selection and the treatment of unsuitable subgrade are engineering decisions requiring a qualified geotechnical or pavement engineer. Lime and cement binders are hazardous materials requiring controls determined by a site-specific risk assessment. Whether a subgrade condition constitutes a latent condition, and whether a non-conformance or a delay in hold point release gives rise to an entitlement, are matters for advice on the specific contract and facts.

Writing a tender? Let’s write it together.

HoursMon–Fri 7am–5pm AEST