Chapter 38 — Heavy Civil Construction: Roads, Bridges, Utilities, and Infrastructure

The Hook: Forty Feet of Rock

Day nineteen at Cottonwood Creek. Ten after seven in the morning, thirty-eight degrees, and the creek is running loud enough that Del Ferraro has to lean in to be heard.

Del Ferraro is the superintendent on the Cottonwood Creek Bridge Replacement — $18.7 million, a state department of transportation as the owner, a unit-price contract, 210 working days on the clock. He has built four bridges. He is standing at the base of a drill rig watching a 36-inch auger that is not going anywhere.

The shafts for Pier 2 are supposed to go to elevation 1,206. Eleven of them, thirty-six inches in diameter, drilled from a working platform at elevation 1,262 down through river alluvium — sand, gravel, some cobble — to tip on competent rock with a nominal four-foot seat. Fifty-six feet of shaft apiece. The geotechnical report and the boring logs bound into the contract documents show competent rock at roughly elevation 1,202 to 1,210 across the pier, which is why the shafts are designed the way they are and why Del's crew planned on a shaft a day.

Shaft 2-1 went fine. Shaft 2-2 went fine. On 2-3, the rig's rate of penetration fell off a cliff at elevation 1,247 — forty-one feet above where the logs said rock started — and the crew spent eleven hours on a shaft that was supposed to take six. Same on 2-4. Same on 2-5. As of this morning, eight of the eleven shafts have hit something the auger will not chew, forty-odd feet high, and Del has burned through the rock tooling he budgeted for the entire job in nineteen days.

"That's the third bit," Del says. "I've got a rig sitting on rock I didn't buy, on a schedule I signed, and a boring log in my truck that says this rock is a story below me."

Standing three feet from the rig, close enough that the DOT's shaft inspector keeps glancing over, is Ingrid Sørensen, the project engineer. She is not arguing with Del. She is holding a folding rule, a camera, and a bound field book with numbered pages, and she is writing.

For every shaft, she is recording four things: the exact elevation at which the drill's rate of penetration changed, the clock time it changed, the tooling on the rig at that moment, and what came out of the hole. She has a five-gallon bucket of spoil labeled 2-5, EL 1244–1248 sitting on the tailgate with a placard in it and a photograph of it beside a tape. She has asked the agency's inspector to initial the elevation entries in her book. He has. She has already sent one email, at 6:52 a.m., to the resident engineer with the subject line NOTICE — Pier 2 drilled shafts — subsurface conditions differing from contract boring logs — request inspection prior to further excavation.

Del is frustrated. Ingrid is calm. Both of them are behaving correctly, and here is why.

On a lump-sum building contract, this morning is a catastrophe. If Kestrel hit unexpected rock forty feet high in the mass excavation at Northgate, we would own it unless we could prove the contract documents affirmatively told us something different — and we would be arguing about it while the schedule bled at $10,650 a calendar day. Lump sum means we took the quantity risk. We priced a hole, and the hole is whatever the hole turns out to be.

On a unit-price contract with a public agency, this morning is not a catastrophe. It is a measurement question and possibly a claim — and the distance between those two outcomes is entirely a documentation problem. The drilled shafts are paid at $412 per linear foot of 36-inch shaft, measured in place. If the shafts get deeper, Del gets paid for more feet, automatically, with no change order at all. But the character of the work changed — rock instead of soil, different tooling, a different production rate, a different cost per foot — and that is not automatically paid. That is a differing site conditions question, and every differing-site-conditions clause I have ever read turns on notice, on the owner's chance to look before you dig, and on contemporaneous records of what was actually there.

Which is why Ingrid is standing in the cold with a field book instead of sitting in the trailer being upset.

By the end of this chapter you will know exactly what she is doing, why the field book beats the memory, how the money gets recovered through two different mechanisms that pay for two different things, and why a building estimator thinks in assemblies while a heavy civil estimator thinks in cycles.

Cottonwood Creek, Del Ferraro, Ingrid Sørensen, Kestrel Construction Group, and every other company, person, and project in this book are Tier-3 illustrative composites. The numbers are internally consistent and realistic. They are not a real project, and no real agency's specification is quoted or paraphrased anywhere in this chapter.

🏃 Fast Track: If you have run unit-price public work before, skim §38.1 and §38.2 and go straight to §38.3 (the quantity variation arithmetic), §38.4 (differing site conditions worked end to end), and §38.5 (mass haul and linear scheduling). §38.11 is the transfer table — read that one even if you read nothing else.

🔬 Deep Dive: Unit-price risk allocation is Chapter 4; clause anatomy is Appendix G. Cycle-time production estimating and fleet matching are Chapter 21. Notice, entitlement, causation, and damages are Chapter 33. Earthwork and structures vocabulary is Chapter 8. Volume and unit conversions are Appendix A.

38.1 What "Heavy Civil" Actually Covers

Heavy civil — also called heavy/highway, infrastructure, or horizontal construction — is the part of the industry that builds the things that are not buildings. The work is linear or site-wide rather than stacked; the product is usually public; the money usually comes out of a government budget; and the dominant cost driver is equipment and its production rate, not trade subcontracts.

Here is the landscape. Every row is a genuinely different business with a different owner culture, and the last column is the one that decides whether you make money.

Two patterns run through the whole table.

First, the owner is usually a public agency. That changes how you get the work (open competitive bidding, price-only award in most cases — Chapter 15), how you get paid (measured quantities against a published schedule), what wages you pay (Chapter 20), and who can eventually look at your records.

Second, the contract is usually unit price — a specific, deliberate allocation of risk that I introduced in Chapter 4 and am now going to make you live inside.

💡 Aha moment. In building work, the general contractor is a coordinator of subcontracts. Something like 80–90% of a commercial GC's contract value is bought out to trades, and the GC's core skill is managing other companies' crews and the interfaces between them. In heavy civil, a prime contractor commonly self-performs the majority of the dollar value — earthwork, structures, drainage, paving — and subcontracts only specialties. So the heavy civil manager's core skill is not procurement. It is production: getting a fleet of machines and a crew to move a measured quantity of material per hour, day after day, at a cost per unit below the price you bid. Everything else in this chapter follows from that sentence.

38.2 The Public Owner and Its Specification

The biggest cultural shock for a building manager walking onto a DOT job is this: the specification was written before your project existed, it applies to every project the agency lets, and it is not negotiable.

A state DOT publishes a book — commonly called the standard specifications — that governs materials, workmanship, measurement, and payment for every item of work the agency buys. Alongside it sits a set of standard drawings (also called standard plans): pre-approved details for guardrail, inlets, headwalls, joints, pavement markings, and dozens of other recurring elements. Your contract then adds special provisions — project-specific language that adds to, deletes from, or amends the standard specifications for this job only.

The order in which those documents govern is written into the contract, and it matters exactly the way order of precedence mattered in Chapter 7. Generally the special provisions modify the standard specifications, and the plans, standard drawings, and specifications interact in a stated hierarchy — but the hierarchy varies by agency and you must read the one in your contract. Do not assume the building-industry order of precedence you learned earlier carries over.

Four consequences follow, and each one costs somebody money every year.

1. You are presumed to know the standard specifications. They are public and free. "I didn't know your spec required a pre-blast survey" is not an argument, it is an admission. Building contractors are used to a project manual assembled for their project. This one was not.

2. Changes are rare, formal, and slow. A public agency cannot casually accept a substitution the way a private owner can. The engineer of record may lack authority to alter a standard detail without a formal process. Budget authority is legislated. This is not obstruction; it is a public agency spending public money under rules designed to prevent favoritism.

3. The inspector's authority is defined in writing. On a DOT job there is a resident engineer or project engineer representing the agency, plus inspectors in the field with a written scope of authority. Their daily records are the agency's official record of your work. What they write down about your quantities is your payment, at least until you dispute it — which is why Ingrid asks the inspector to initial her elevations rather than keeping a private record and hoping.

4. Every pay item has a written measurement and payment clause. This is the one that separates money from wishes.

38.2.1 Measurement and payment — read this clause before you price the item

For each pay item, the specification states three things:

Three examples of how much the method matters:

• Neat-line measurement. A structural excavation item is often measured to a neat line — a theoretical prism defined by the plans (say, the footing footprint plus two feet, on a defined slope) — regardless of how much you actually dug. If you over-excavate for working room, or because the bank sloughed, you moved that dirt for free. Your unit price has to carry it.
• Plan quantity versus measured quantity. Some items are paid at plan quantity: the number in the bid schedule is final, no field measurement, unless the plans themselves change. Others are paid at measured quantity, computed from field cross-sections or field measurement. On a plan-quantity item, an error in the plans becomes your windfall or your loss depending on direction, and there is no re-measure to save you.
• Weighed versus computed. Asphalt paid by the TON is weighed on a certified scale with tickets. Asphalt paid by the SY at a specified thickness is computed geometrically. The same paving, paid two different ways, rewards completely opposite field behavior. Paid by the ton, an extra quarter-inch of thickness is revenue. Paid by the square yard, that same quarter-inch is material you gave away.

⚖️ What the contract says. The most expensive words in a DOT specification are some version of "…shall be considered incidental to the contract unit price for…". Incidental means required, inspected, enforceable, and not separately paid. Dewatering is frequently incidental to excavation. Trench shoring is very frequently incidental to pipe. Clearing inside the pay limits is often incidental to earthwork. Test rolling, cleanup along the alignment, and disposal of surplus material are commonly incidental. Temporary erosion control may be paid or incidental depending on the agency. Every one of those is real work with real cost. If you did not put it into a unit price, you have already lost the money and you will not find out for three months. So read the measurement and payment paragraph of every significant item before you price it. Not the item description in the bid schedule — the clause.

🏗️ From the field. Ingrid's first job out of school was a small municipal drainage contract. The bid schedule had an item for 18-inch storm pipe at a linear-foot price and a separate item for structural excavation. She priced the pipe at a decent number and assumed the trench came out of the excavation item. It did not — the excavation item on that contract applied only to structures, and the measurement clause for pipe said trench excavation, bedding, backfill, and compaction were incidental to the pipe. The company ate roughly $61,000 on 4,200 linear feet of pipe: about $14.50 a foot of unpaid trench work. Nobody was fired. The estimating checklist grew a line that reads, in capital letters, READ THE MEASUREMENT CLAUSE, NOT THE ITEM NAME.

🔄 Check your understanding. A bid schedule shows "Structural Excavation, 4,800 CY, $38.50/CY." Before you use that number, what three things must you find in the specification, and what does each one change?

38.3 Unit-Price Contracting, in Depth

Go back to Chapter 4 for a second and recall the one-sentence version: in a unit-price contract, the owner takes the quantity risk and the contractor takes the productivity risk. That is the deal. Now let us live inside it.

38.3.1 The bid schedule, and the thing everyone gets wrong

A DOT bid package includes a bid schedule: a numbered list of pay items, each with a description, a unit of measure, and an engineer's estimated quantity. You fill in a unit price for each item. The agency multiplies your rate by its estimated quantity, adds up the extensions, and that sum is your bid total — the number the award is based on.

Here is the thing everyone gets wrong on their first unit-price job:

You are not bidding the total. You are bidding a rate. The total is arithmetic the agency does to compare bidders. What you will actually be paid is your rate times the quantity that is measured in the field.

The estimated quantities are the designer's calculation. They are usually close. They are not guaranteed, and they are not a representation that you will build that much. If the plans call for 4,800 CY of structural excavation and the field cross-sections come to 5,410 CY, you are paid for 5,410 CY at your rate. No change order. No negotiation. It is not a change to the contract at all — it is simply the contract working the way it was written.

That is a genuinely lovely property. It also contains a trap.

38.3.2 The trap: fixed costs hide inside a variable rate

Every unit price is built from two very different kinds of cost.

You bid a single rate per unit. So the fixed cost has to be divided by the estimated quantity and buried inside the rate. The moment the actual quantity differs from the estimate, your fixed-cost recovery breaks.

💰 Money check — Cottonwood Creek structural excavation at $38.50/CY.

Here is how Kestrel's estimator built that canonical rate. The engineer's estimated quantity for structural excavation on the bridge was 4,800 CY.

At the estimated quantity, the item performs exactly as planned:

• Revenue: 4,800 CY × $38.50 = $184,800
• Cost: (4,800 × $20.00) + $48,000 = $96,000 + $48,000 = $144,000
• Margin: $40,800, which is $8.50/CY — precisely what was bid.

Now suppose the actual measured quantity comes in at 60% of estimate — 2,880 CY — because the designer's assumed excavation limits were conservative and the resident engineer authorizes tighter neat lines.

• Revenue: 2,880 CY × $38.50 = $110,880
• Cost: (2,880 × $20.00) + $48,000 = $57,600 + $48,000 = $105,600
• Margin: $5,280, which is $1.83/CY.

The margin fell by $35,520 — 87% of it — on an item where nothing went wrong in the field. Break the loss into its two halves so you can see the mechanism:

What it means for the job: on a unit-price contract, an item can destroy your margin by not happening. Nobody was slow, nothing broke, no rain fell. The quantity simply came in low, and the fixed cost stayed exactly where it was.

🔍 Why this works. The fixed cost did not care about the quantity, but the payment mechanism only pays by quantity. You built a recovery plan on the assumption of 4,800 units and then sold 2,880 of them. This is the same arithmetic that makes an underloaded factory unprofitable, and it is the same arithmetic behind extended general conditions on Northgate: $2,900,000 of general conditions divided by 565 calendar days gives $5,150/CD, and if the job runs longer the daily rate is what you claim, because the total was never a function of production in the first place. Fixed cost divided by an assumed denominator is one of the most reliable ways to lose money in this industry, and it shows up in five different disguises across this book.

38.3.3 The quantity variation clause — why it exists and what it does

Public agencies know about the arithmetic above. That is why nearly every unit-price contract contains a quantity variation clause (sometimes called a variation in estimated quantities or significant change in character of work clause).

The typical structure: if the actual quantity of a major item varies from the estimated quantity by more than a stated threshold — commonly on the order of ±25%, but this number, the definition of "major item," and the mechanics all vary by agency and must be verified in the contract you signed — then either party may request an adjustment to the unit price. The adjustment usually applies only to the varied portion (the quantity above or below the threshold), not to the whole item.

Note the clause is bilateral, and understand why. It protects the contractor when a quantity collapses and fixed costs strand. It equally protects the agency when a quantity balloons: if the excavation had come in at 9,600 CY instead of 4,800 CY, the contractor would collect $10.00/CY of fixed-cost recovery on 4,800 CY of extra work whose fixed cost was already paid for — a windfall of $48,000 for doing nothing new. The clause lets the agency ask for the rate to come down on the excess.

Applying it to our example: the quantity came in 40% below estimate, which is beyond a 25% threshold, so the clause is available. What can Kestrel realistically recover?

⚖️ What the contract says. Three things to know before you count on this clause. (1) Most clauses of this type compensate for stranded cost, not for lost anticipated profit on work never performed. The $16,320 of profit on the 1,920 CY that never happened is usually not recoverable at all. (2) Many clauses apply only to increases, or only to decreases, or have different mechanics for each — read yours. (3) Many require a written request within a stated period, sometimes before the item is complete. A contractor who discovers a 40% underrun at final measurement, six months after the item finished, may have a valid grievance and no remedy. Thresholds, definitions, and notice periods vary by agency and change over time. Verify them in your contract; do not carry a number from your last job.

38.3.4 Unbalanced bidding — the technique, the arithmetic, and the line

Because you are bidding rates rather than a total, you have a degree of freedom the lump-sum bidder does not have: you can move money between items without changing your bid total. That is unbalanced bidding, and it comes in two flavors that are ethically and legally very different.

Flavor one: front-end loading

Load early items (mobilization, clearing, excavation) above cost and shave late items (paving, striping, landscaping) below, so the money arrives sooner. Same total, better cash flow.

Identical bid. But $160,000 arrives with the first or second pay application instead of eight months later. If your cost of money is 9% a year, that is roughly $160,000 × 0.09 × (8 ÷ 12) ≈ $9,600 of carrying cost avoided, plus reduced draw on a line of credit — which, as Chapter 32 argued, is the currency contractors actually run out of.

Agencies know this too. Many cap the mobilization pay item at a stated percentage of the contract, or pay it in tranches tied to percent complete, precisely to limit it. Modest front-end loading that reflects genuine early cost is normal, expected, and defensible. Loading far beyond your real early cost is an interest-free loan from a public agency that never agreed to make one.

Flavor two: quantity-error unbalancing — the one that bites

This is different in kind. Here you are not moving money in time; you are betting against the engineer's estimated quantities. You raise the rate on an item you believe is understated and drop the rate on an item you believe is overstated, keeping the bid total identical.

Take the same job. Kestrel's estimator does his own takeoff and concludes roadway excavation will actually run about 78,000 CY, not 62,000, and that the borrow item is badly overstated: the plans show 24,000 CY of borrow, but with the extra cut he expects to need only about 8,000 CY.

The bid total moves by eighty dollars. Now watch what happens when the actual quantities show up:

Two hundred twenty-nine thousand dollars, on a bid that was arithmetically identical at award. That is why agencies scrutinize bid schedules for unbalancing and why a bid found to be materially unbalanced — meaning the unbalancing creates a reasonable doubt that the agency will end up paying the lowest ultimate cost — can be rejected as non-responsive. Federal-aid work and most state programs have specific policies on this; they differ by agency, and the review is done by people who compare your rates against every other bidder's and against the agency's own historical price database.

Here is the line, and I want to be plain about it.

Building a rate that reflects your cost, your production rate, and your fixed-cost recovery is estimating. Two contractors will produce different rates for the same item, and that is the market working. Loading the front end to cover the genuine cost of mobilizing is estimating. Pricing an item deliberately away from its cost because you believe the agency's quantity is wrong is not estimating. It is a wager, and the counterparty is the public.

At the far end of that spectrum sits something worse. If you know the plans contain a quantity error — because your survey found it, or you built the adjacent phase — and you price to exploit it while staying silent, you have moved from aggressive bidding toward misrepresentation. Bid protests and false-claims exposure live down that road, and the specifics vary by jurisdiction and by whether federal funds are involved. Bring that question to a construction attorney, not to a bid-day argument at 10:45 a.m.

And here is the practical argument, for anyone unmoved by the ethical one: unbalancing amplifies your error in both directions. In the table above, if the estimator is wrong — if the excavation is really 62,000 CY and the borrow really is 24,000 CY — the two schedules pay exactly the same. But if he is wrong the other way, if excavation runs short and borrow runs long, the unbalanced schedule loses just as fast as it would have won. You have converted a construction business into a position on somebody else's arithmetic. Tomás Reyes, Kestrel's chief estimator, puts it this way: "An unbalanced bid is a number that stopped being a story about risk and started being a story about luck."

38.3.5 Force account — getting paid for work that has no pay item

When the agency directs work that does not fit any pay item, and no rate can be agreed in advance, the contract's force account provision (sometimes called extra work at agreed rates, or simply extra work) governs. It is the unit-price world's cost-plus mechanism: you perform the work and are paid documented cost plus stated percentages for overhead and profit, with equipment paid at rates defined by the specification.

Force account pays well and is a nightmare to administer, because the payment is the paperwork.

That last line is the whole game. An unsigned force account ticket is a piece of paper with your opinion on it. A signed one is, in practice, a payment. The rule Del gives every new engineer: no ticket leaves the site unsigned, and if the inspector will not sign it, you write on the ticket that he declined and why, and you send it to the resident engineer that afternoon.

That is the same principle that cost Kestrel $43,650 on CO #14 at Northgate — the change that was built before it was priced, with no time-and-material tickets for the first four days. Of $186,400 actually incurred, only $121,000 could be substantiated, and the settlement landed at $142,750. The price of a change is set by what you can document, not by what it cost you. (Chapter 31 works that one in full.)

38.3.6 Build the bid schedule

📋 Try it — price a six-item unit-price bid, then get paid on it.

You are the estimator for the approach roadway package on Cottonwood Creek. The agency's bid schedule and your own cost build-up are below. Traffic control, trench shoring, dewatering, and surface restoration are all incidental under this contract and are already inside the variable costs shown.

The bid schedule (engineer's estimated quantities):

Your cost build-up:

Answer these:

(a) Build a balanced set of six unit prices — spreading the $148,000 of job overhead across the measured items in proportion to their variable cost, then applying markup. (b) What is your total bid? (c) The job finishes. Roadway excavation came in 40% below the estimated quantity; Class A concrete came in 15% above. Everything else measured as estimated. What are you paid, and what happened to your margin? (d) Assume the contract's quantity variation clause uses a ±25% threshold and applies to major items (defined as any item exceeding 10% of the total contract amount). Is the clause triggered on any item? If so, what would you actually ask for? (e) Sketch an unbalanced version of this bid, and explain why the agency would reject it.

🧩 Productive struggle. Before you read §38.4, sit with this one for four minutes.

The Cottonwood Creek Pier 2 drilled shafts are bid as 616 LF of 36-inch drilled shaft at $412/LF, measured in place — eleven shafts, each drilled from a working platform at elevation 1,262 down to a tip at elevation 1,206, which is fifty-six feet apiece. That extension is $253,792.

Del has just found competent rock at elevation 1,247 on eight of the eleven shafts — forty-one feet above where the logs put it. The agency's engineer will almost certainly confirm the rock and raise the tip elevation on those eight shafts, because a shaft bearing on good rock does not need to keep going.

Two questions, and write your answers down before you read on:

1. What happens to the paid quantity?
2. Rock drilling costs Kestrel roughly six to seven times per foot what augering alluvium costs. So: does the unit-price contract protect Del here, hurt him, or both? Name the specific mechanism you would reach for.

38.4 Differing Site Conditions — the Defining Risk of Heavy Civil

Here is the answer to the productive struggle, and it is the most important paragraph in this chapter.

The paid quantity goes down. The engineer cores the material, confirms competent rock at elevation 1,247, and revises the tip on the eight affected shafts to elevation 1,243 — the same nominal four-foot seat, just forty-one feet higher. Each of those shafts is now 19 LF instead of 56 LF.

So the unit-price mechanism, which people describe as protecting the contractor, has just cut Del's revenue on this item nearly in half — while his cost per foot went up by a factor of six on every foot of rock he had to cut. The contract paid him less for work that cost him more.

💡 Aha moment. Unit price does not protect you from bad ground. It protects you from quantity error in the direction the drawings anticipated. When the ground itself is different in character from what the documents portrayed, the unit-price schedule is silent, or worse, actively works against you. That is precisely the gap the differing site conditions clause exists to fill, and it is why heavy civil managers can recite their clause from memory.

38.4.1 Type I and Type II

Nearly every public construction contract in the United States, and many private ones, contains a differing site conditions clause (often abbreviated DSC). The near-universal structure recognizes two categories:

The practical guidance: always analyze Type I first. Go find what the contract documents indicated, in writing, and show the difference. Type II is the argument you make when the documents indicated nothing at all, and it is a much steeper hill because "unusual" is an argument, not a measurement.

38.4.2 What the clause requires of you — and the way most valid claims die

A typical differing site conditions clause imposes conditions on the contractor before the agency owes anything. The specifics vary by contract, so read yours — but the shape is remarkably consistent:

⚖️ What the contract says. I want to state this as bluntly as I know how, because I have watched it happen four times in twenty-two years and it never stops being avoidable:

The most common way a valid differing-site-conditions claim is lost is that the contractor excavated the evidence before anyone looked at it.

Not because the claim was weak. Not because entitlement was absent. Because a superintendent who was behind schedule made the entirely rational field decision to keep working, and by the time the letter went out the condition was in a truck, on a stockpile, four miles away, and unverifiable. The agency's answer — "we were never given the opportunity to inspect" — is usually enough, and in some contracts it is an express bar. The clause is not a formality. It is a condition precedent, which is legal language for the door only opens if you knock in the specified way, and whether a particular notice provision is enforced strictly varies by jurisdiction and by contract. Ask your attorney about yours before you need it, not after.

This is why Ingrid's 6:52 a.m. email exists and why the rig sat while the resident engineer drove out. The three hours the rig lost that morning were the cheapest three hours on the job.

🔍 Why this works. Think about the notice requirement from the owner's side and it stops feeling like a trap. A public agency's engineer has options the moment she learns about the rock: verify it and revise the design (which is what happened — and saved the agency $121,952 of shaft), redirect the work, delete it, adjust the sequence, or accept it and price it. Every one of those options requires that the condition still exist. Notice is not a paperwork ritual designed to defeat contractors. It is the mechanism that keeps the owner's options alive, and the reason it is strict is that a late notice destroys options that were worth real money. Understand that and you will write better notices, because you will write them to preserve the owner's choices rather than to preserve your grievance.

38.4.3 Boring logs, disclaimers, and what a bidder may rely on

Every contractor who bids heavy civil eventually has this argument.

The contract includes a geotechnical report with boring logs — a record of what a drill found at specific locations, with sample descriptions, blow counts, groundwater observations, and often laboratory results. It also includes, almost always, a disclaimer saying in some form that the subsurface information is furnished for information only, that it represents conditions only at the boring locations on the dates drilled, that the agency does not warrant it, and that the bidder is responsible for its own interpretation and encouraged to make its own investigation.

So which is it? Is the log a representation you can rely on, or a document that disclaims itself?

The honest answer, and I am going to hedge here because it deserves hedging: courts and boards in many jurisdictions have held that a general disclaimer does not automatically defeat a Type I claim where the contract documents affirmatively indicated a condition and the bidder reasonably relied on that indication. The reasoning usually runs that an owner who furnishes information invites reliance, and that a broad disclaimer cannot convert affirmative data into nothing. But this varies significantly by jurisdiction, by the specific language, and by the facts — how strong the indication was, how reasonable the reliance was, whether a site investigation would have revealed the truth, and whether the contractor complied with the clause. Some contracts contain far more aggressive language than the general disclaimer described here. Do not read a paragraph in a textbook and conclude you are covered. Read your clause, and when it matters, ask a construction attorney in your jurisdiction.

Two practical rules follow that are true regardless of jurisdiction:

1. What the documents indicate is broader than what they state. A profile drawn showing a rock line, a note about anticipated excavation classification, a pay item structured a particular way — all of these can be indications, even if no sentence promises anything. Ingrid's first job on this claim was to build a one-page exhibit showing the boring logs, the plan profile, and the actual measured elevations on the same vertical scale. That exhibit did more work than any letter.
2. Do the site investigation the contract tells you to do, and document that you did it. A contractor who walked the site, reviewed the logs, and made a reasonable interpretation stands very differently from one who did neither.

38.4.4 The geotechnical baseline report

There is a better mechanism, and it comes out of tunneling and heavy underground work.

A geotechnical baseline report (GBR) is a contract document that does not merely describe what borings found. It states the conditions the parties agree to treat as the contractual baseline — for example, that the contractor shall be deemed to have priced a stated range of rock strength, a stated percentage of a particular ground class along the alignment, or a stated groundwater inflow. Conditions worse than the baseline are a differing site condition and are compensable. Conditions better than the baseline belong to the contractor as an opportunity, or in some structures are shared.

Why this is genuinely good practice, and why it is spreading beyond tunneling:

That last row is the one to internalize, and it is theme #1 of this book in its purest form: risk that cannot be priced gets priced anyway, badly, by everyone. A GBR does not eliminate ground risk. It converts an unpriceable unknown into an allocated, measurable one. It requires real geotechnical investment up front by the owner, and it requires an owner willing to admit that it, not the contractor, is the party that can most cheaply reduce uncertainty about its own site.

38.4.5 Working the Cottonwood Creek rock event — which mechanism pays for what

Now put it all together. This is the analysis Ingrid ran, on one page, eleven days after the notice. The single most important idea here is that three different contract mechanisms were live simultaneously and each one paid for something different. A manager who reaches for only one of them leaves money on the table or, worse, pursues the wrong argument.

The build-up of the differing-site-conditions ask. Kestrel's $412/LF rate was built on augering alluvium at about 8 LF per hour with a rig-and-crew cost of roughly $1,680 per hour — a variable cost of $210/LF — plus $52,000 of item fixed cost spread across 616 LF ($84.42/LF), marked up about 40%. In rock, the same rig made about 1.2 LF per hour: a variable cost of $1,400/LF.

And the honest outcome, negotiated over five weeks:

And the scoreboard for the whole item, honestly:

What it means for the job: Kestrel recovered the overwhelming majority of its cost and less than half of its margin. That is what a well-documented, well-noticed, well-analyzed differing-site-conditions event looks like when it goes well. Nobody gets made whole. You get made close.

Time. The event pushed Pier 2 by 9 working days. Kestrel requested 9; the agency granted 6. Two of the requested days coincided with a high-water event that would have stopped in-stream work regardless — genuine concurrency — and one day was waiting on a tooling delivery, which is Kestrel's own supply chain. Six working days on a 210 working day contract, granted in writing, with the shaft activity's float recalculated in the update. That is the right outcome and Del argued about it for a week before he agreed.

🔄 Check your understanding. Ingrid's notice went out at 6:52 a.m. on the day the crew stopped drilling, before anything was disturbed. Name three specific things that email preserved that would have been unavailable if she had sent it eleven days later, after all eight shafts were poured.

38.5 Estimating and Planning as Production, Not Assemblies

Here is the sentence that organizes this whole section:

A building estimator thinks in assemblies. A heavy civil estimator thinks in cycles.

When Tomás Reyes prices interior partitions at Northgate, he thinks in assemblies: a linear foot of 3⅝-inch metal stud partition, 16 inches on center, gypsum board both sides, taped and finished, is worth some number of dollars per linear foot, and 18,600 LF of it is worth that number times 18,600. The assembly is a recipe. Its cost per unit is roughly stable because the work is the same work each time.

When Del prices structural excavation, there is no recipe. There is a machine, a cycle time, a payload, and an hourly cost. The unit cost is not a property of the work; it is the output of a production calculation, and it changes completely if the haul is a mile longer, the bucket is half a yard smaller, or the excavator has to wait on trucks.

This is the direct continuation of Chapter 21, and if the phrase fleet matching is fuzzy, go back and reread it now.

38.5.1 A worked production estimate — where $20.00/CY came from

Let us build the variable half of the Cottonwood Creek structural excavation rate from first principles.

Step 1 — What the machine produces per cycle.

A 1.5 CY excavator working inside a braced excavation next to the abutment. Bucket fill factor for damp granular material in confined conditions: 0.85.

1.5 CY × 0.85 = 1.275 loose cubic yards (LCY) per pass

Step 2 — How many cycles per hour.

Cycle time in a confined pit — short swing, but constant interference from sheeting, dewatering hose, and a survey check every few passes: 34 seconds.

3,600 s ÷ 34 s = 105.9 passes/hr × 1.275 LCY = 135.0 LCY/hr at 100% efficiency

Step 3 — Nobody works 60 minutes in an hour.

Job efficiency of 45 productive minutes per hour (0.75) — realistic for structural excavation with an inspector present, not for open mass excavation, which runs higher.

135.0 × 0.75 = 101.3 LCY/hr

Step 4 — Convert loose to bank, because you are paid in bank measure.

Swell for this material: 28%.

101.3 LCY ÷ 1.28 = 79.1 bank cubic yards (BCY) per hour

Step 5 — Match the trucks so the excavator never waits. (This is the step people skip, and it is where fleets lose 20% of their production.)

Trucks required = truck cycle ÷ load time = 23.23 ÷ 6.23 = 3.73 → use 4 trucks

Check: 4 trucks × (60 ÷ 23.23) trips/hr × 14 LCY = 144.6 LCY/hr at 100%, or 108.5 LCY/hr at 0.75 efficiency. That exceeds the excavator's 101.3 LCY/hr, so the excavator governs and never waits on a truck. Correct fleet. With three trucks, production drops to 81.4 LCY/hr and you have paid for an excavator that idles 20% of the day.

Step 6 — Hourly cost of the whole spread.

Step 7 — Cost per bank cubic yard moved.

$988/hr ÷ 79.1 BCY/hr = $12.49 per BCY moved

Plus disposal at the permitted spoil site, $2.30/BCY → $14.79 per BCY moved

Step 8 — The step that separates heavy civil from everything else: you do not get paid for everything you move.

This item is measured to a neat line. The actual excavated volume, including safe sloping, working room outside the neat prism, and the sloughing you will absolutely get in this material, runs about 1.35 times the neat-line volume.

$14.79 per BCY moved × 1.35 BCY moved per BCY paid = $19.97 per PAID CY

Round to $20.00/CY — the variable cost inside the $38.50 bid rate from §38.3.2.

What it means for the job: the number $20.00 is not a price you looked up. It is 35% overburden, four trucks, a 34-second cycle, and a 3.2-mile haul. Change the haul to 6 miles and the truck cycle goes to about 31 minutes, you need 5 trucks, and the paid unit cost climbs past $24/CY — a 20% swing on a single logistics assumption. On a heavy civil bid, the haul road is a line item in disguise.

🔄 Check your understanding. The agency's plans permit spoil disposal on-site in a designated waste area 0.4 miles away instead of the permitted off-site facility 3.2 miles away, and the $2.30/BCY tip fee disappears. Roughly what happens to the paid unit cost, and — more importantly — what do you check before you re-price the item?

38.5.2 Mass haul — balancing dirt along a line

On a building, earthwork is a volume problem: 44,000 CY cut, 12,000 CY fill, 32,000 CY net export at Northgate, and the only real question is where the trucks go.

On a linear project, earthwork is a distribution problem. You have cut in some places and fill in others, spread over miles, and the money is not in digging — it is in how far each cubic yard travels and how many times you touch it.

The tool is the mass haul diagram. It is simple once you see it: walk the alignment station by station, and keep a running algebraic total of cut (positive) minus fill (negative), adjusted for shrinkage or swell so the volumes are comparable. Plot that running total against station.

Here is a mass haul table for the Cottonwood Creek approach roadway, stations 10+00 through 30+00:

📊 Diagram (described) — the mass haul curve. Plot the cumulative column against station. The curve rises through every cut reach and falls through every fill reach. The vertical axis is cubic yards; the horizontal axis is distance along the road.

   CUMULATIVE
   VOLUME (CY)
  +16,000 |                   #                        <- peak: max cut before fill starts
          |
  +12,000 |             #           #
          |
   +8,000 |                                     #
          |
   +4,000 |       #                       #  ← BALANCE LINE (haul limit) ─ ─ ─ ─
          |                                  #
        0 |#- - - - - - - - - - - - #- - - - - - - - - - - - - - - - - -#- - -
          |                              #
   -4,000 |                            #
          +--+-----+-----+-----+-----+-----+-----+-----+-----+-----+-----+
           10+00 12+00 14+00 16+00 18+00 20+00 22+00 24+00 26+00 28+00 30+00
                        STATION  (100-ft stations)

   RISING segment  = you are in CUT      (material available)
   FALLING segment = you are in FILL     (material needed)
   PEAK  (16+00)   = cut ends, fill begins — this is where haul direction reverses
   VALLEY (24+00)  = fill ends, cut begins again
   Curve ENDS at +1,700  → 1,700 CY SURPLUS to waste
   Curve DIPS to −2,100  → 2,100 CY DEFICIT: you must BORROW in that reach

How to read it, in four rules that take five minutes to learn and pay for the rest of your career:

1. Rising means cut, falling means fill. The slope is the intensity of the operation.
2. Any horizontal line you draw across the curve is a balance line. Between two points where a horizontal line crosses the curve, the cut exactly equals the fill. That reach can be balanced within itself — the dirt never has to leave.
3. The area between the curve and the balance line is the haul — volume times distance, in station-yards. Area is what you actually pay for. Two projects with identical cut and fill quantities can have wildly different costs because one has a fat curve and one has a thin one.
4. Where the curve ends above zero, you have surplus to waste. Where it dips below zero, you have a deficit to borrow. Here: 1,700 CY of waste at the end and a 2,100 CY borrow requirement around station 22+00 to 24+00 — and those are two separate pay items with two separate prices, not one net number.

Free haul and overhaul. Traditionally, the excavation unit price includes moving material up to a stated free haul distance — the distance beyond which additional payment applies. Material moved farther earns overhaul, historically paid in units like station-yards or cubic yard-miles beyond the free haul limit. Many modern agencies have dropped separate overhaul items entirely and simply require the contractor to include all haul in the excavation price, which shifts the entire distribution risk onto the bidder. Which convention your contract uses is one of the first things to determine, because it changes your excavation rate substantially and it varies by agency.

💰 Money check — why the curve is worth reading before you bid. Suppose the designer's intended earthwork plan hauls the 5,400 CY of cut at station 26+00 backward to fill station 22+00 — about 400 feet of haul. But a bidder who reads the curve notices that the fill at 20+00 (7,400 CY) is more efficiently fed from the 15,700 CY peak at 16+00, and that the 26+00 cut is better used at 30+00. Rebalancing shortens the average haul on roughly 12,000 CY by an average of 900 feet. At a scraper-and-truck haul cost of about $0.55 per CY per 100 feet of one-way haul, that is:

12,000 CY × 9 stations × $0.55 = $59,400

on a job where the entire earthwork item might bid at $600,000. That is a 10% swing that lives entirely in a diagram most building estimators have never drawn.

38.5.3 Linear scheduling — why a highway is scheduled by station, not by activity

In Chapter 14 you built a critical path method (CPM) schedule: activities, logic, durations, a forward pass, a backward pass, float, and a critical path. That technique is correct and it is the right tool for Northgate, where a thousand different activities happen in a bounded footprint.

It is the wrong primary tool for a road.

On a linear project, you do not have a thousand different activities. You have five or six crews doing the same thing over and over, moving along an alignment. The clearing crew moves down the road. The earthwork crew follows. Drainage follows that. Subgrade and base follow drainage. Paving follows base. Every crew does one kind of work, repeatedly, at a rate — feet per day, stations per week.

Recall the Chapter 37 lesson from Harbor Ridge: the unit of production changes the scheduling technique. Colton Reyes could not schedule eleven houses with a critical path because his unit of production was a repeating house, so he used line of balance and managed a rate. Del cannot schedule a road with a critical path for the same reason in a different geometry: his unit of production is a repeating station, so he uses a time-location chart (also called a linear schedule, a time-distance diagram, or location-based scheduling).

📊 Diagram (described) — the time-location chart. Horizontal axis: station along the alignment. Vertical axis: working day, running downward. Each crew is a line, not a bar. The line's slope is the crew's production rate: a steep line (nearly vertical) is a slow crew; a flat line (nearly horizontal) is a fast one. Two lines that converge or cross mean two crews trying to occupy the same station at the same time — which on a 40-foot-wide road corridor is not a coordination problem, it is a physical impossibility and a safety hazard.

 WD |        COTTONWOOD CREEK APPROACH ROADWAY — TIME-LOCATION CHART
----+-------------------------------------------------------
  0 |  A
  5 |              A
 10 |                           A
 15 |     B                                 A
 20 |          B                                         A
 25 |               B
 30 |  C                B
 35 |         C               B
 40 |                C             B
 45 |                       C           B
 50 |                           C            B
 55 |  D                        C                 B
 60 |            D              C                      B
 65 |                      D    C     <<< CONFLICT ZONE
 70 |                                D  C   <<< CONFLICT ZONE
 75 |                                          D   C
 80 |            E                                       D
 85 |                                     E
 88 |                                                    E
----+-------------------------------------------------------
     10+00 12+00 14+00 16+00 18+00 20+00 22+00 24+00 26+00 28+00 30+00
                          S T A T I O N

 A = Clearing & grubbing   B = Earthwork/grading   C = Drainage & utilities
 D = Subgrade & aggregate base    E = Paving

Read the chart and three things jump out that no bar chart would have told you.

First, crew C flattens out — it stops. Between working day 48 and working day 66, the drainage line does not move: it sits at station 20+00 for eighteen working days. That is the rock trench reach. On a Gantt chart this appears as "Drainage — WD 30 to WD 78, 48 days." One bar. Nothing about it says the crew is going to sit still in one place for eighteen days.

Second, crews C and D collide. The base crew (D) is moving at 0.4 stations per day. It reaches station 20+00 around working day 67 — one and a half days after the drainage crew finally clears it, and then the two lines run almost on top of each other from station 20+00 to station 30+00 for the rest of the job. Two crews, one corridor, ten stations, twelve working days of near-contact. That is the single most important fact about this schedule, and a CPM bar chart hides it completely, because on a bar chart "Drainage" and "Base" simply overlap in time — which is normal, expected, and looks fine.

Third, the fix is visible on the chart itself. You can see three options without running a single calculation:

🔍 Why this works. A CPM network answers "what must finish before what?" A time-location chart answers "who is standing where, when?" Those are different questions, and on a linear job the second one is the one that costs you money. The classic linear failure is not a missed predecessor — it is two crews, a pipe crew and a grading crew, in a 40-foot corridor, taking turns waiting on each other for three weeks while both charge full crew cost to the job. CPM cannot see it because CPM has no concept of space. The time-location chart's whole content is space. Most serious heavy civil contractors run both: a CPM for contract time, milestones, and any delay analysis (the agency will want CPM), and a linear schedule for actually running the work.

⚠️ Safety alert. A converging-crew conflict on a linear chart is a safety finding, not just a productivity finding. Two crews in the same corridor means a trench crew working next to loaded haul trucks, a grade checker in the swing radius of an excavator, and a pipe layer in a trench beside a vibratory roller. This is exactly theme #4 in geometry: schedule pressure is a hazard. When you see two lines converge on that chart, the first phone call is to the superintendent about the plan, and the second is to the safety manager about what happens if the plan does not change.

38.5.4 Seasons, weather, and the working-day contract

Northgate's contract time is 565 calendar days, and if Kestrel slips substantial completion the liquidated damages run at $5,500 per calendar day whether it is raining, snowing, or beautiful. Cottonwood Creek's contract time is 210 working days. That difference is not a formatting choice. It is a different allocation of weather risk, and it exists because heavy civil work genuinely cannot proceed in certain conditions.

How working days actually work. The mechanics vary by agency, but the shape is consistent and you must learn yours:

1. The contract states the number of working days — 210 at Cottonwood Creek.
2. The specification defines what makes a day chargeable: typically a day on which the contractor could have worked some stated portion (often a majority) of a normal day on the controlling item of work.
3. The resident engineer makes that determination daily and issues a periodic statement — often weekly or monthly — showing days charged.
4. The contractor has a short, stated window to object in writing to a day charged. Miss it and you may have waived the objection.
5. Days not charged: weather days meeting the criteria, agency-caused suspensions, holidays, and days outside a permitted work season.

⚖️ What the contract says. That fourth item is the one that costs contractors real money every year. If your project engineer does not reconcile the agency's day charges against your own daily reports every single week, you will arrive at the end of the job having silently accepted charged days you could have contested — and by then the objection window closed months ago. Ingrid reconciles Cottonwood Creek's day charges every Monday morning against her daily reports and the on-site rain gauge, and she has contested eleven days so far and won seven. Seven working days on a 210-day contract is 3.3% of the schedule, recovered with a spreadsheet and a habit. This is Chapter 26 applied with a stopwatch, and it is why the daily report is not busywork.

And notice the incentive shift. On a calendar-day contract, weather risk sits with the contractor, so you price a weather contingency into the bid — every bidder does, and the owner pays for it whether or not it rains. On a working-day contract, the agency absorbs weather-day risk in time while the contractor still absorbs it in cost (your equipment and salaried staff are on the job whether or not the day is charged). Neither is free. Neither is universally better. They are two different answers to who owns the weather, which is theme #1 wearing a raincoat.

38.6 The Work Itself, at a Manager's Level

You do not need to be able to run a drill rig to manage one. You do need to know what each operation is, what governs its production, what it costs when it goes wrong, and which of its risks belong to you. What follows is that briefing — not a means-and-methods manual. For the underlying construction vocabulary, Chapter 8 is the reference.

38.6.1 Earthwork and grading at scale — and rock

Mass grading on a linear job is the operation the mass haul diagram describes, executed with scrapers, articulated trucks, excavators, dozers, and compactors, controlled increasingly by machine control (GPS or robotic total station guidance that puts the design surface in the operator's cab and largely eliminates grade stakes — see Chapter 39).

The manager's daily questions are always the same three: Is the fleet balanced? Is the moisture right? Is the haul road maintained? A neglected haul road is the most common quiet killer of earthwork production — cycle times creep up 15% over two weeks and nobody notices because no single day is bad.

Rock excavation is where the money and the arguments live. Two things matter to you:

1. Classification. Does your contract have a separate pay item for rock excavation, or is excavation unclassified — one price for whatever you find? Unclassified excavation moves the entire rock risk to the contractor, which is why the differing site conditions clause matters so much more on an unclassified contract. Check this before you price a cubic yard.
2. How you break it. Ripping with a large dozer, hammering with a hydraulic breaker, or blasting. Production and cost differ by an order of magnitude between them.

Blasting carries an administrative burden most building managers have never seen:

🏗️ From the field. The pre-blast survey is the cheapest insurance in heavy civil and the most commonly shortened. A contractor I know skipped four houses at the edge of the survey radius because the owners were not home on the two days the survey crew was there. Eleven weeks later, three of those four owners filed damage claims. The two the surveyor had documented were resolved in a week — the cracks were in the photos, dated, before the first shot. The unsurveyed ones took fourteen months and cost, between settlement and legal fees, something north of six figures. Nobody won an argument about physics. The photos won.

38.6.2 Drainage and underground utilities — and the most common source of delay in the industry

Pipe work looks simple and is not. A storm or sanitary installation is a repeating cycle: locate, saw-cut, excavate, shore, dewater, prepare the bedding, lay and joint the pipe, backfill in lifts, compact, test, and restore the surface. Production is governed by depth, groundwater, soil, traffic control, and the number of times per day you have to stop.

The parts a manager must get right:

• Bedding and backfill are structural, not cleanup. A flexible pipe gets most of its strength from the compacted material around it. Skimp on haunching under the pipe springline and you get deflection failures that show up on the video inspection at the end of the job, when the pipe is 14 feet deep under a finished road.
• Trench safety is not optional and not incidental to good intentions. OSHA's excavation standard (29 CFR 1926 Subpart P) requires a protective system — sloping, benching, shielding (trench box), or shoring — for excavations at and beyond the threshold depth it specifies, requires a competent person to inspect daily and after every rainstorm, and governs spoil pile setback, access and egress, and atmospheric hazards. Read Chapter 24 and Appendix F.
• Dewatering — sumps, well points, or deep wells — is frequently incidental to the pipe or excavation item, and it is frequently the largest single unbudgeted cost on a wet job. Also check the discharge: you usually need a permit and a treatment or settling requirement to put groundwater into a stream.
• Trenchless methods exist because open-cut is impossible or unacceptable in some places:

Trenchless is far more expensive per linear foot than open-cut and is often much cheaper overall once you price the traffic control, restoration, and business disruption that open-cut would require. That comparison is a management decision, not a technical one, and it belongs in the estimate.

⚠️ Safety alert. Trench collapse is among the most lethal hazards in construction. A cubic yard of soil weighs roughly as much as a small car, a collapse gives no warning, and the second fatality in a trench event is very often a coworker who jumped in to help. Three non-negotiables: a protective system appropriate to the depth and soil, a competent person inspecting daily and after every rain event, and a rescue rule that everyone on the crew has heard out loud — nobody enters an unprotected trench to perform a rescue. Call for trained rescue. This is the practice that turns one fatality into one fatality instead of two.

Utility conflicts — the number one delay generator on urban civil work

Now the thing that will actually cause your delay. On urban and suburban civil work, conflicts with existing underground utilities are, by wide consensus in the industry, the most common cause of delay and disruption. Not weather. Not the agency. Pipes and cables that are not where the drawings say they are.

Here is the structural problem in one paragraph. The as-built records of buried utilities in most jurisdictions were created over decades by many different owners, at varying standards, often by field crews measuring from features that no longer exist, and sometimes never updated when the utility was relocated. They are frequently wrong — in horizontal position, in depth, in size, in material, and occasionally in existence. A one-call locate service marks what the utility owners' records and locating equipment can find, which is not the same as what is there. Abandoned facilities are notoriously undocumented.

And here is the contractual problem: the utility owner is usually not a party to your contract. The gas company that has to relocate a main in your way has no contract with you, no liquidated damages, no schedule obligation to you, and its own capital budget and crew backlog. You cannot direct it. You often cannot even effectively escalate to it. Your remedy, if you have one, runs against the owner through the delay provisions of your contract — and many public contracts contain language limiting the agency's liability for utility delays, or granting time but not money. Read that clause before you bid an urban utility job.

The four practices that actually reduce this exposure:

🏗️ From the field. Del's rule, which he learned expensively: "Pothole every crossing on the critical reach two weeks ahead of the pipe crew. It costs about four hundred dollars a hole. A crew standing still costs about six thousand dollars a day. Do the arithmetic once and then stop arguing with me about it." On a job with 22 crossings, that is $8,800 of potholing against an exposure of a day and a half of stopped crew per undiscovered conflict. If potholing prevents two stoppages, it paid for itself twice.

38.6.3 Structures — foundations, substructure, superstructure

A bridge is built from the ground up in a strict sequence, and each step has a different production logic.

Three management points that matter more than the technique:

Concrete supply continuity is a schedule risk with a specification behind it. A bridge deck placement is a continuous operation. Once you start, you cannot stop without creating a cold joint the specification does not permit. That means the batch plant, the truck fleet, the pump, the finishing machine, the crew, and the weather must all hold for the entire placement — and it means you need a written contingency: a backup plant, a backup pump, and a defined point of no return after which you commit. The reinforcing steel at $1.42/LB and the Class A concrete at $685/CY are the least interesting numbers in that operation.

Rebar quantity is measured by weight, and weight is calculated, not weighed. On a unit-price contract, reinforcing steel is typically paid by the pound based on computed weight from the placing drawings — not the weight of steel you actually bought. Waste, laps beyond those shown, and support steel are usually incidental. On a bridge with $1.42/LB steel and a heavily congested pier cap, the difference between computed and purchased weight is real money that lives entirely inside your unit price.

Formwork on a bridge is a temporary structure with engineering behind it. Falsework supporting a cast-in-place span is a designed, submitted, often independently reviewed structure — see Chapter 22. Agencies typically require sealed falsework drawings and a stated inspection sequence before you place a yard of concrete.

38.6.4 Bridge construction methods

💡 Aha moment. ABC and slide-in exist for a reason that has nothing to do with construction cost and everything to do with user cost. If 42,000 vehicles a day detour eight miles for six months, the aggregate cost to the traveling public — fuel, time, crash exposure — can dwarf the construction cost of the bridge. Agencies have started to price that. So a method that costs 25% more to build and closes the road for a weekend instead of six months can be the cheaper project by the measure the owner actually cares about. When you cannot explain why an owner chose an expensive method, look for a cost that is not on your estimate.

38.6.5 Paving — the production system, and pay factors

Paving is the most production-like operation in construction. It is also the operation where quality is measured statistically and paid statistically, which is a concept building contractors find genuinely alien.

The layers. Subgrade (the prepared native or embankment surface, proof-rolled and accepted), aggregate base (graded stone, placed and compacted to a density specification), then either asphalt courses or concrete pavement.

The asphalt paving train is a single production system, and its throughput is set by its slowest element:

  PLANT ──► HAUL TRUCKS ──► MTV ──► PAVER ──► BREAKDOWN ──► INTERMEDIATE ──► FINISH
  (tons/hr)   (cycle time)  (surge)  (fpm)      ROLLER          ROLLER        ROLLER
      │            │           │        │           │                │            │
      └────────────┴───────────┴────────┴───────────┴────────────────┴────────────┘
                 THE SLOWEST ELEMENT SETS THE PRODUCTION OF ALL OF THEM
                 A paver that stops and restarts leaves a mark you cannot roll out.

The governing principle: the paver must never stop. A stopped paver creates a transverse joint, a density anomaly, and a smoothness defect all at once. So the entire system — plant output, truck count, haul cycle, and the material transfer vehicle (MTV) that keeps a surge of mix in front of the paver — exists to keep one machine moving at a constant speed. This is fleet matching again, exactly as in §38.5.1, with the paver in the excavator's role.

Mix design and temperature. The mix is designed and approved before you pave; you cannot change it in the field. Temperature governs compaction: asphalt must be compacted while it is above a minimum temperature, and the available compaction time shrinks with thin lifts, cold ambient air, and wind. This is why the paving season is real and why the last lift often waits for spring.

Concrete pavement is a different production system — a slipform paver, a continuous concrete supply, dowel baskets or a dowel bar inserter at transverse joints, tining or grooving for texture, curing compound, and sawing joints in a window that opens when the slab can hold a saw and closes before it cracks on its own. The sawing window is measured in hours and moves with the weather.

Pay factors — getting paid a percentage of your unit price

Here is the concept. On many acceptance items — asphalt density, thickness, air voids, concrete strength, pavement smoothness — the agency does not simply accept or reject. It samples statistically by lot, computes a statistic, and applies a pay factor: a multiplier on the contract unit price for that lot.

• Pay factor 1.00 — full payment; the lot met the specification.
• Pay factor below 1.00 — the lot is acceptable but deficient; you are paid a reduced price for material that stays in place.
• Pay factor above 1.00 — some agencies pay an incentive for results better than the target, particularly on smoothness.
• Below a floor, the lot is removed and replaced, or accepted at a steep reduction at the engineer's option.

Pay factor schedules, statistics, lot sizes, and limits vary substantially by agency and by item. The schedule below is illustrative only, constructed to show the arithmetic — it is not any agency's table and must not be used as one.

💰 Money check — one deficient asphalt lot.

Illustrative density pay schedule, acceptance by lot on mean in-place density as a percent of maximum theoretical density:

Lot 7 is 1,180 tons of surface course at a contract unit price of $93.37/TON. The agency's randomly located cores average 91.6%.

Full-price value: 1,180 TON × $93.37 = $110,176.60 Pay factor: 91.6% falls in the 91.0–91.9 band → 0.95 Payment: $110,176.60 × 0.95 = $104,667.77 Revenue lost: $5,508.83

What it means for the job: you bought the same aggregate, ran the same plant, paid the same crew, and burned the same fuel. Every dollar of that $5,509 comes out of margin. On this item you planned about $12.18/TON of overhead and profit (§38.3.6), so 1,180 tons was carrying about $14,372 of margin. One deficient lot just consumed 38% of the margin on that lot — because two roller passes were skipped at the end of a long day, or the mat cooled faster than the crew expected.

That is why serious paving contractors run their own density gauge behind the rollers in real time rather than waiting for the agency's cores. The gauge costs a fraction of one deficient lot.

⚖️ What the contract says — and the line you do not cross. Acceptance sampling on a DOT job is normally located by a random process controlled by the agency, precisely so that neither party can choose favorable locations. Influencing where a core is taken, working the mat differently in the area you expect to be tested, or any arrangement with an inspector about sample location is not aggressive quality management. It is falsification of the acceptance record on a publicly funded project, and on federal-aid work it can carry consequences far beyond the contract — debarment, and in serious cases criminal exposure. The specifics vary by jurisdiction and by funding source. The rule does not: you improve a pay factor by rolling the mat, not by managing the test. If your density is short, add a roller, tighten the pattern, shorten the haul, or slow the paver. Those are the only four answers.

🔄 Check your understanding. Your paving superintendent proposes running the mat 0.15 inches thicker than the plan thickness "so we have margin on density and smoothness." The item is paid by the TON. Then a second job comes along where the same item is paid by the SY at plan thickness. What is your answer on each job, and what changes?

38.7 Traffic Control: The Public Is Inside Your Site

At Northgate, the site is fenced. Every person inside has been oriented, wears personal protective equipment, and is there on purpose. Access is controlled by a badge and a gate. That is the mental model most building managers carry, and on a road job it is completely wrong.

On a highway project, the public drives through the middle of your site at fifty-five miles an hour, all day, every day, having received no orientation and signed nothing. Some of them are distracted. Some are impaired. Some are angry about the delay. They are inside your work area and you cannot exclude them, because the entire point of the project is that the road stays open.

Everything about work-zone management follows from that sentence.

The traffic control plan (TCP) — sometimes called a temporary traffic control or maintenance-of-traffic plan — is a contract document showing how traffic will be handled through each phase of construction: signing, channelizing devices, tapers, buffer spaces, lane shifts, detours, flagging operations, and pedestrian and bicycle accommodation. Agencies base these on standardized national and state guidance and typically require them to be prepared or reviewed by a qualified person. Your TCP is not a suggestion; it is enforceable, and deviating from it without approval is a violation of the contract and, more to the point, a way for someone to be killed.

The core structure of any work zone, worth committing to memory:

   ADVANCE WARNING AREA  ►  TRANSITION AREA  ►  BUFFER  ►  WORK AREA  ►  TERMINATION
   (signs tell drivers      (the taper moves   (empty!    (crews and    (traffic
    what is coming)          traffic over)      no crew,   equipment)     resumes
                                                no gear)                 normal)
   ───────────────────────────────────────────────────────────────────────────────►
                              DIRECTION OF TRAVEL

The buffer space is the part people give away when they are behind schedule. It is empty on purpose: it is the distance an errant vehicle travels before it reaches a worker. Storing materials, parking a pickup, or letting the crew creep upstream into the buffer is the most common work-zone violation and the one with the most direct line to a fatality.

Lane closure windows are contractual. Many urban contracts permit closures only during off-peak hours — say 9 p.m. to 5 a.m. — and impose lane rental or per-minute liquidated damages for reopening late. Those charges can be brutal and are meant to be: they price the delay the public suffers. If your contract has a lane rental provision, the schedule for a night's work is not a plan, it is a countdown, and you rehearse it.

⚠️ Safety alert. Work-zone safety data consistently show that a substantial share of worker fatalities in highway work zones involve a vehicle intruding into the work space — a motorist entering the area where crews are working — rather than an incident caused by construction equipment. The practical implications are unglamorous and they save lives: positive protection (barrier, not cones) wherever the exposure and duration justify it; high-visibility apparel on every person in the right of way, without exception; a designated spotter when equipment backs near traffic; internal traffic control for your own trucks so workers on foot and moving equipment are separated; law-enforcement presence where speeds justify it; and a stated rule that nobody works with their back to live traffic. Every one of those is cheap. A funeral is not.

38.8 Environmental Compliance and Permitting

Heavy civil work happens in creeks, wetlands, floodplains, and habitat. The regulatory framework is real, enforced, and capable of stopping your job for a season. I am going to describe these by function, because the specific statutes, agencies, permit types, and thresholds vary by country, state, and locality, and change over time. Never carry a permit assumption from your last job. Read this project's permits, and read the conditions attached to them — the conditions are where the schedule constraints hide.

⚖️ What the contract says. Two questions decide how much of this you own. First: who obtained the permits? On most DOT work the agency obtains the major environmental permits before letting the contract, and the contractor is obligated to comply with their conditions — which means you must read them, because they contain sequencing requirements your CPM must reflect. On some contracts the contractor obtains certain permits itself, and that is a different risk entirely. Second: what happens when a condition changes or an unanticipated resource is discovered? Archaeological discovery in particular is usually treated as a compensable suspension or a differing site condition — but the mechanism, the notice, and the compensation vary by contract. Find the clause on day one, not on the day the operator uncovers something.

🏗️ From the field. The most expensive environmental event I have watched was not a violation. It was a bald calendar mistake. A contractor sequenced a culvert replacement to begin two weeks after the in-stream work window closed, because the earthwork ahead of it ran long and nobody had drawn the window on the schedule as a hard constraint. The window closed. The culvert waited eleven months. The contractor kept a small crew, a field office, and a bond in place across a winter for a $340,000 item. Nothing was cited. Nothing was violated. The line on the schedule simply was not there. Put every permit window on your schedule as a date constraint with a name, and review them at every monthly update.

38.9 Safety: A Different Risk Profile

The hazards on a heavy civil job are not the hazards on a building job, and a safety program transplanted without adjustment will protect against the wrong things.

Bea Salgado's argument from Chapter 24 applies here without modification, and maybe more forcefully: safety is a property of the production system, not a rulebook. The scaffold near-miss at Northgate in week 34 had three findings, and the one that mattered was the third — a crew running behind after the steel acceleration, carrying an unwritten "make it up" pressure. On a heavy civil job, the equivalent finding is a pipe crew ten stations behind the linear schedule with a base crew breathing down its neck, in a trench, next to live traffic. The time-location chart in §38.5.3 is a safety document. When you see two lines converge, you are looking at the conditions that produce an incident, four weeks before it happens, which is exactly where this book keeps telling you the job is won.

38.10 Prevailing Wage, DBE, and Public Accountability

Public money carries public obligations. Three of them shape how you staff and administer a heavy civil job.

Prevailing wage. On federal and federally assisted construction, the Davis-Bacon Act and related statutes generally require payment of locally prevailing wages and fringe benefits by craft classification, determined and published by the federal government. Most states have their own prevailing-wage statutes covering state and local public work, commonly called "little Davis-Bacon" acts — and these vary enormously: some states have none, some have thresholds, some have different classification systems, and the law changes. Practical consequences, which Chapter 20 develops in full:

• Your labor rate is set by classification and by the determination applicable to the contract, not by what you would otherwise pay.
• Certified payroll must be submitted at a stated frequency, with worker classification, hours, rates, deductions, and a signed statement of compliance.
• Classification is where the risk lives. Paying a laborer's rate to someone performing operator work is a violation whether or not it was intentional. So is misclassifying an employee as an independent contractor. Underpayment findings carry back wages, penalties, and potential debarment — and debarment ends a public-work contractor's business.
• Apprentice ratios, fringe benefit crediting, and the treatment of owner-operators are technical and easy to get wrong. Get them right at buyout, not at audit.

DBE participation. Federally assisted transportation contracts typically carry Disadvantaged Business Enterprise requirements: a program goal, often a contract-specific goal, and an obligation either to meet it or to document good faith efforts to do so. Two things to understand as a manager:

1. The commitment is enforceable. A DBE commitment made in your bid becomes a contract obligation, and substituting or terminating a committed DBE firm generally requires the agency's written consent.
2. The firm must perform a commercially useful function. A DBE that is a pass-through — one whose scope is actually performed and managed by someone else — does not count, and arrangements structured to create the appearance of participation are fraud. This is not a gray area, and it has ended companies.

Public accountability generally. Bids are opened publicly and read aloud; the tabulation is public. Your contract, your change orders, and often your correspondence may be subject to public-records requests. Complaints about your job go to elected officials. Prompt-payment statutes governing how quickly you must pay subcontractors after being paid vary by state and are actively enforced on public work. And on federally funded work there are additional layers — Buy America requirements on certain materials, wage and civil rights compliance, and specific reporting obligations. All of this varies by jurisdiction, program, and funding source. Verify what applies to your contract; do not reason from your last one.

💡 Aha moment. Every one of these requirements is administrative cost, and administrative cost is a fixed cost, which sends you straight back to §38.3.2. A compliance manager, a payroll clerk, and the hours your project engineer spends on certified payroll and DBE reporting are the same kind of cost as the $52,000 of drill rig mobilization: they do not scale with quantity, and they must be recovered inside unit prices built on estimated quantities. This is a genuine and underappreciated reason small contractors struggle on public work — not that they cannot do the work, but that the fixed administrative load per dollar of contract is higher for them, and unit-price bidding hides that cost inside rates that must then compete against a larger firm's.

38.11 What Transfers From Building Work, and What Does Not

If you are moving from building construction to heavy civil, this table is the honest inventory. I have watched good building managers struggle in civil for exactly the reasons in the right column.

🔄 Check your understanding. Name the three specific things a building project manager is most likely to get wrong in her first month on a DOT job, and the concrete habit that prevents each.

Spaced Review

Close the book on this chapter for a moment and answer these from memory. Write the answers down — retrieval you have to work for is retrieval that sticks.

From Chapter 37. Colton Reyes told me at Harbor Ridge that he had "no critical path and eleven buildings," and that what he had instead was a rate. What was the underlying principle there — the one that explains why he used line of balance and Del uses a time-location chart while Wei Chen uses CPM on Northgate?

From Chapter 21. Before you look: what does it mean for a fleet to be matched, and what is the number you compute to decide how many trucks to put behind an excavator?

Deep callback to Chapter 4. In one sentence each: in a unit-price contract, who owns quantity risk and who owns productivity risk? Then the harder half — name the clause that exists specifically because that clean division breaks down, and say what breaks.

Project Checkpoint: The Willow Street Heavy-Civil Contrast Memo

In Chapter 37 you wrote the residential contrast memo — how you would run Willow Street if it were 34 houses instead of one building. This checkpoint is its mirror image: re-cast the Willow Street site work as a unit-price heavy civil contract and see what changes.

Your scope for this exercise is the horizontal work only: the site utilities (including relocating that existing 8-inch water main), the parking and drive paving, and the site earthwork on the 2.1-acre site. Assume the City of Rivermont Parks & Recreation lets this scope as a separate unit-price contract instead of rolling it into the $6.8M lump sum.

Produce five things.

1. A bid schedule. Eight to twelve pay items covering the earthwork, utilities, and paving. For each: item description, unit of measure, estimated quantity from your Appendix K quantity data, and — this is the part that teaches — a one-line measurement method ("measured in place by cross-section, neat line"; "measured horizontally along the centerline of pipe, no deduction for structures"; "measured by weight from certified scale tickets"). Getting the measurement method written down is worth more than getting the quantity exactly right.

2. Unit prices built from production, not assemblies. Pick your three largest items and build each rate the way §38.5.1 built $20.00/CY: machine, cycle time, payload, efficiency, swell where it applies, crew and equipment hourly cost, then fixed cost divided by estimated quantity, then markup. Show the arithmetic. Do not look up a unit cost — derive one. Then compare your derived rate to the assembly-based number you used in your Chapter 13 detailed estimate and explain any gap.

3. An incidental-work list. For every item on your schedule, state what is incidental to it under your assumed measurement clauses: trench shoring, dewatering, bedding, backfill, compaction, testing, traffic control, erosion control, restoration, surplus disposal. Then total the cost of everything you just declared incidental. That number is what you would have lost by pricing item names instead of clauses.

4. Field measures to protect quantity. Write five specific things you would do differently in the field on a unit-price contract that you would not bother with on lump sum. Suggested starting points: daily quantity logs signed by the inspector; joint measurement of trench depths before backfill; photographs with a scale and a station marker at every measured event; force account tickets signed the day worked; weekly reconciliation of your quantities against the agency's.

5. A differing-site-conditions exposure comparison. One page. On the lump-sum version, where does your risk sit if you hit rock or groundwater or find that the 8-inch main is three feet off from the record drawing? On the unit-price version, which mechanism responds — the bid schedule, the quantity variation clause, the DSC clause, or force account — and what notice does each require? Be specific about what you would do in the first hour of discovering the main in the wrong place.

Next chapter you will build the technology adoption plan for Willow Street — and item 4 of this memo is going to feed straight into it, because most of what makes quantity measurement reliable in modern heavy civil is a drone, a GPS rover, and a photo log with coordinates attached.

Chapter Summary

Heavy civil is not building work with dirt instead of drywall. It is a different commercial system, and this table is the reference card for it.

The decision framework, when something unexpected comes out of the ground:

1. Stop. Do not disturb the condition.
2. Notify in writing, today, and ask the agency to inspect before the condition is disturbed.
3. Document: elevation, station, photographs with scale, samples, production records, and the inspector's initials.
4. Classify: is this a quantity change (the bid schedule handles it automatically), a character change (differing site conditions), directed extra work (force account), or a quantity variation beyond the threshold (the variation clause)? It is frequently more than one, and each pays for something different.
5. Segregate the cost from day one, in its own cost code.
6. Ask for the time separately from the money, in working days, in writing.

What's Next

Chapter 39 picks up exactly where this chapter's measurement problem leaves off. Almost every hard thing in heavy civil — knowing where the utility actually is, measuring a quantity you can defend, documenting a condition before you disturb it, keeping a paving train at constant speed, proving a subgrade was built to grade — is being changed right now by drones, reality capture, machine control, and GPS measurement. You have spent this chapter learning why the record is worth money. Next you will look at the tools that make the record cheap, what they actually cost, and which of them return more than they cost on a job the size of yours.