Chapter 15 — Construction Monitoring and Field Testing

QA versus QC: Who Is Responsible for What

ASCE Manual 73 (ASCE 2012) separates two functions that are often blurred together as “QA/QC.” Quality assurance is the planned, systematic work of the owner or its agent to gain confidence that the design complies with codes and criteria and that construction complies with the contract documents; it verifies that the other parties’ quality control is working. Quality control belongs to the design professional (control of the contract documents) and to the constructor (control of the work), and includes observations, calculations, inspections, tests and documentation.

The owner sets the value placed on quality, funds it, and decides whether to use a standing generic QA plan, a project-specific plan, or a tailored generic plan. The owner may specify the designer’s and constructor’s QC plans prescriptively or in performance terms. During construction the owner must choose who inspects: the owner, the design professional, an independent agency, a code official, or the constructor. The Manual stresses that none is always best and the choice should balance cost, safety, schedule and quality. A resident project representative (RPR) may oversee the constructor’s QC and reviews the constructor’s QC plan before production starts.

A typical constructor project-specific QC plan addresses a skilled workforce, a QC organization, a submittal schedule, inspections, mock-ups and sample construction, a testing plan including special inspections, documentation, and corrective action when acceptance criteria are not met. It should also cover subcontractor qualification, control of measuring and test equipment, and segregation of nonconforming material. The Manual summarizes responsibilities in a chart: the owner establishes the QA program; the designer develops design-phase QC; the constructor develops and documents construction QC; and all communicate regularly.

For railroad work: On a railroad-funded embankment or subgrade project the railroad (or its engineering consultant acting as RPR) performs QA, and the contractor performs QC testing such as compaction and gradation. When the geotechnical consultant’s field staff perform acceptance testing, the scope should say whether that testing is QA (owner verification) or is replacing contractor QC, because the two carry different responsibility for rework and different testing frequencies. State this explicitly in the construction monitoring scope and the GSP.

Source: American Society of Civil Engineers (2012). Quality in the Constructed Project: A Guide for Owners, Designers, and Constructors, 3rd ed. ASCE Manuals and Reports on Engineering Practice No. 73. Reston, VA: ASCE, pp. 169–170, 191–199. Summarized and paraphrased. Full reference under Supplemental Sources.

Specifying, Measuring and Paying for Driven Piles

Goldbloom gives a checklist of what a bearing-pile section must state. Pile monitoring depends on these items, so they should be confirmed before driving starts:

• the design capacity;
• a requirement that excavation be complete before piles are driven;
• driving criteria: minimum tip elevation, minimum penetration, and final blow count in blows per inch over the last 6 in. Where test piles and load tests are included, set the blow count after the load tests;
• driving sequence for groups, from the center outward;
• a minimum distance from concrete less than 7 days old (60 ft in his example);
• submittal of the driving equipment;
• limits on jetting, which must stop 3 to 6 ft above the expected final tip so the last few feet are driven;
• conditions under which predrilling is allowed;
• whether cut-offs may be spliced and reused.

His example tolerance is a maximum deviation of ¼ in per ft from vertical or plan batter, and 3 in in any direction at cut-off elevation.

For payment, he separates fixed costs from variable costs so that neither party gains or loses unfairly when the pile footage changes:

• Mobilization of pile-driving equipment: lump sum, including moves between locations and demobilization. This is the largest fixed cost, especially for barge work.
• Production piles: per linear foot from tip to actual cut-off, by type and size. The unit price normally covers jetting, followers, rejected and damaged piles, and, depending on the contract, cut-offs.
• Cut-offs on timber or precast piles: pay separately by the linear foot when the engineer sets the ordered lengths.
• Test piles: pay separately. They cost more per foot than production piles even if they become permanent.
• Load tests: pay per each, because the number is uncertain.
• Predrilled holes: per linear foot. Pipe-pile points, H-pile tips and splices: per each.

For railroad work: On bridge and culvert replacement projects, the pile monitoring plan in this section should check that the driving criteria and tolerances in the contract match the design, and that the pay items match what the inspector will record. Tip elevation, penetration, final set, cut-off elevation, and splice and cut-off lengths should all appear on the pile log. Where tie-in work is restricted to short track windows, mobilization and remobilization between windows should be priced explicitly so they do not end up as claims.

Source: Goldbloom, J. (1989). Engineering Construction Specifications: The Road to Better Quality, Lower Cost, Reduced Litigation. New York: Van Nostrand Reinhold, pp. 20, 225–233. Summarized and paraphrased. Full reference under Supplemental Sources.

Recording Piling Data to Match the Pay Items

Seeley shows how CESMM3 (Classes P and Q) splits piling so that a variation can be valued fairly. Items are grouped by piles of the same type, material and section at one location. Each group produces two or three items:

1. The number of piles. This covers moving the rig to each position, setting up and preparing to drive or bore.
2. The length of piles. This covers the material. For cast-in-place piles it is measured from the cut-off level to the toe level that was expressly required.
3. The depth driven or bored, measured from a stated commencing surface. Raked piles are identified, with the rake stated.

Interlocking sheet piling is measured by area: the mean horizontal length of the wall times the depth. Corner, junction, closure and taper piles are extra linear items. Plant mobilization and temporary staging suit fixed method-related charges, and operating the piling plant suits a time-related charge. Ancillary work is measured only when expressly required. That includes cutting off surplus length, preparing pile heads, and extensions, which are paid as a number item for each joint plus a length item split at about 10 ft. Breaking out obstructions above the founding stratum is paid by the hour, and only when ordered, so the engineer can control and record it. Load tests and non-destructive tests are separate items.

For railroad work: on bridge pier, abutment and retaining-wall piling, the inspector’s driving record is the evidence for each of these quantities, so record them on the log as the work proceeds. For every pile, record:

• pile ID and group
• ground or commencing elevation when driving starts
• tip elevation reached and the cut-off elevation
• length delivered, length cut off and splices made
• batter
• start and stop times of any ordered obstruction work, with the reason
• any re-strike or dynamic test

When a pile runs past its plan length, separate records for the pile count, the length furnished and the depth driven let the change be priced from the bid items. It does not have to become a lump-sum claim. Where piling is done under track windows or next to live track, note the time lost waiting for trains, so that time-related standby can be reconciled against the railroad’s own records.

Source: Seeley, I. H. (1993). Civil Engineering Quantities, 5th ed. Basingstoke: The Macmillan Press, pp. 166–167. Summarized and paraphrased. Full reference under Supplemental Sources.

Tracking Cost Against Plan and Managing Change in Construction-Phase Services

During execution, the course compares progress to plan with three tools: quality reviews, Gantt schedule performance charts and cost performance charts. Its cost chart plots cumulative planned value (PV) against cumulative actual cost (AC) by week. In the example, planned value reaches $44,000 by week 5 against an actual cost of $66,000, so spending is 50% above plan and the gap has widened every week.

Table 15-PM1. Cumulative planned value and actual cost, $1,000 (after Skill Right 2013, Cost Performance Chart)

For change, the course separates customer-requested changes (usually the largest source) from internal requests, regulation and team-generated changes. Its procedure is: hold a team meeting, explain the change, get feedback, identify alternatives, prepare a decision matrix, select a recommendation, present it to management or the customer, and implement the approved course.

For railroad work: a PV-versus-AC chart cannot by itself tell overspending apart from work being done ahead of schedule. Add the budgeted value of the work actually performed, measured in physical units: inspector shifts actually needed, density tests, piles observed, monitoring rounds and reports issued. Construction-phase geotechnical budgets on railroad projects are often consumed by standby and repeat mobilizations when track windows are cancelled or the contractor’s sequence changes. Log those days as a separate cost line from the first occurrence. Then, when the change procedure is invoked, the decision matrix compares real alternatives:

• keep a full-time inspector on standby;
• switch to on-call coverage with notice requirements;
• request a scope and fee amendment for the extended duration.

Report cost status with the same frequency and in the same document as the QA/QC test summaries, so the railroad sees quality, schedule and cost together.

Source: Skill Right (2013). Project Management Tools and Techniques: An Introductory Course in Project Management Foundations [training slides sponsored by Texas Instruments, Inc.; minor modifications by N. Santiago, Spring 2013 version], slides 81–83, 86–87. Summarized and paraphrased; figures reproduced for internal reference. Full reference under Supplemental Sources.

Change Orders, Claims and the Daily Job Log

Sears et al. list the main sources of project change: added or altered work ordered by the owner or engineer, contractor value-engineering proposals, delays caused by the owner or other contractors, and job-site conditions that differ materially from those anticipated. Their procedure for handling a change:

1. Document each change when it arises, estimate its cost and time effects before the work starts, and get written approval to proceed.
2. To find the time effect, bring the schedule network fully up to date without the change, calculate it, then insert the change and recalculate. The difference is the change. A change on a non-critical path may justify no time extension, but it still uses up float.
3. Watch for knock-on effects. A foundation problem can push later work into bad weather. In their bridge example, a delayed abutment could put the work into the season when the dry riverbed becomes a stream.
4. Treat overhead carefully. A fixed markup (10–15% of direct cost is common) covers processing the change but not extended field overhead. Where the duration grows, the network shows how many days of job overhead apply ($869 per day on the example bridge).

Their rule for claims is documentation, and the core record is the daily job log. Each entry records the date, weather, workers by craft, equipment by type, activities started and finished with quantities, subcontractors on site, deliveries (including late or damaged ones), visitors and inspectors, meetings and decisions, problems, and corrective steps. To be accepted as evidence, entries must be original, made on the date shown in the regular course of business, based on the writer’s own knowledge, and kept in a bound book or an equivalently secure electronic system. Entries should state facts without opinion or emotion.

For railroad work: The inspector’s daily report on a railroad job should add the items that drive most disputes there: the track window granted versus planned, with start and stop times; flagging availability; train delays and slow orders in effect; and quantities of undercut, unsuitable material removed and fill placed by station. When the subgrade exposed in an undercut differs from the boring logs, photograph it, note the stations and depths, and notify the railroad’s representative the same day. Then run the “with and without change” schedule comparison before the next window is booked.

Source: Sears, S. K., Sears, G. A., Clough, R. H., Rounds, J. L., and Segner, R. O., Jr. (2015). Construction Project Management: A Practical Guide to Field Construction Management, 6th ed. Hoboken, NJ: John Wiley & Sons, pp. 328–332. Summarized and paraphrased. Full reference under Supplemental Sources.

Daily Logs, Photographs and Event Records as Claim-Grade Documentation

Levy (2007) treats documentation as part of the work itself and places his documentation chapter immediately before the one on claims for that reason. Good records are not volume; they are prompt, accurate entries made when important events happen. For any significant event he recommends recording what happened, when and how it happened, when it was discovered, whether responsibility can be assigned with certainty, who was notified and when, its impact, the immediate and longer-term actions required and by whom, and which party will resolve it. Facts should be separated from opinions; useful remarks at meetings can be noted separately to show intent.

The superintendent’s daily record is the backbone. Levy notes that a bound diary with daily entries meets the legal definition of a business record, while loose sheets may not, and suggests reviewing electronic logs with counsel. At a minimum each entry should give the date (with “No Work” on idle days); weather and temperature at start, midday and end of shift; each subcontractor on site with crew size, work and location; the firm’s own workforce and tasks; visitors and their purpose; inspections by agencies, testing labs or the engineer; work performed and where; deliveries with ticket numbers; unusual events and stoppages; and every accident, reportable or not.

Photographs document progress or lack of it, conditions that differ materially from those expected, support for change orders, and complex details. Levy recommends a tape or rule in close-ups for scale, and specifically lists rock, groundwater and buried debris as conditions to photograph even when they seem unlikely to matter contractually. His closing advice is simply: document, document, document.

For railroad work: The geotechnical field representative’s daily report on a subgrade rehabilitation, slide repair or culvert replacement should add railroad-specific fields: track and work-window times granted and released, the employee in charge or flagman, trains passed during the window, any slow order in effect, and lost window time with its cause. Photograph undercut subgrade, seepage, slip surfaces and unsuitable material in place with a scale and a station reference before it is removed or covered, and link each photo to the daily report and test locations.

Source: Levy, S. M. (2007). Project Management in Construction, 5th ed. New York: McGraw-Hill, pp. 229–230, 255–260. Summarized and paraphrased; figures reproduced for internal reference. Full reference under Supplemental Sources.

Method versus End-Result Requirements and the QC/QA Split

Goldbloom states a basic rule: responsibility for the result follows whoever prescribed the method. If a specification dictates how the work is done and the contractor follows it, the contractor is not liable for a bad result. His common bad example is embankment specifications that fix loose-lift thickness, compactor type and size, and number of passes, and then also demand a minimum in-place density. When the procedure is followed and the density is missed, the owner has no position. His recommended end-result approach is to:

• limit loose-lift thickness and the moisture range;
• specify the required density as a percentage of the laboratory maximum dry density;
• let the contractor choose hauling, spreading, moisture control and compaction equipment;
• have the engineer verify compliance by in-place density tests.

Lift thickness, moisture, the dumping method, and the type and size of the compactor all affect the number of passes needed. That is why the means should be left to the contractor.

He also separates two functions:

• Quality control (QC) is the contractor’s responsibility for running its operations so they produce conforming work.
• Quality assurance (QA) is the owner’s representative checking that the work conforms.

Each technical section should say which tests the engineer will perform and by what method, so the contractor knows how acceptance will be judged. His cautionary example is a contract that made the contractor hire the testing laboratory for in-place density, batch-plant inspection and concrete cylinders. A lab paid by the contractor owes its first loyalty to the contractor, so this cannot serve as QA.

On tolerances, he notes that rigid wording such as “true to line and grade” leads to claims. In one spillway case the court allowed the contractor to recover the cost of forming to tolerances tighter than the finished structure required.

For railroad work: Ballast, subballast and subgrade QA/QC documentation should state for each item whether it is an end-result or a method requirement, and should not mix the two. When a method requirement is unavoidable, such as track-time-limited proof rolling or prescribed lift placement under a work window, write the acceptance criterion as conformance to that method. Keep acceptance testing (nuclear gauge, plate load, proof roll observation) with the railroad’s representative or its own consultant. Contractor testing should be identified as QC only. Give grade and cross-slope tolerances that track and drainage design can actually live with.

Source: Goldbloom, J. (1989). Engineering Construction Specifications: The Road to Better Quality, Lower Cost, Reduced Litigation. New York: Van Nostrand Reinhold, pp. 20–21, 183–184, 188. Summarized and paraphrased. Full reference under Supplemental Sources.

Documenting a Differing Site Condition in the Field

Anglin works through the general-conditions clause on concealed or unforeseen conditions, quoting AIA A201 (Article 4.3.4), and illustrates it with a case from his own practice. The clause recognizes two types:

1. Subsurface or concealed physical conditions that differ materially from those shown in the contract documents.
2. Unknown conditions of an unusual nature that differ materially from those normally encountered in that kind of work.

Whoever observes the condition must give notice promptly, before it is disturbed, and in any case within 21 days. The architect or engineer then investigates. If the condition does differ materially, a change follows; if not, the reasons must be given in writing. Anglin stresses giving written notice of cost or delay even before the amount is known. A continuing delay needs to be documented only once.

In his fire-college case, excavation for a new building uncovered a fire line that turned unexpectedly under the footprint. The contractor took these steps:

• Stopped excavating at the pipe.
• Traced its path on the ground and photographed it.
• Sketched the planned and actual alignments.
• Notified the architect the same day.
• Priced the relocation.
• Issued a notice of continuing delay.

The owner argued that a plan note required the contractor to field-verify utilities. The contractor answered that, under the contract, studying the documents is meant to facilitate construction, not to discover errors. He also showed that no one had specified a particular sequence of excavation. The owner paid.

For railroad work: on subgrade, slope and culvert jobs, the geotechnical field representative is usually the first to see the condition. Typical examples are:

• buried timber or old trestle bents
• cinders, or pockets of fouled ballast deeper than the planned undercut
• springs
• soft clay below the design subgrade elevation
• an unmarked utility or fiber-optic line

The daily report should record:

• the location by milepost and station, offset and depth
• photos with a scale and a reference point
• samples and a quick classification
• a direct comparison with the baseline in the geotechnical report and the boring logs
• the time work stopped and the time it resumed
• who was notified, and when

Railroad and agency contracts set their own notice periods and procedures, so put them in the field instructions before work starts.

Source: Anglin, G. (2020). Introduction to Estimating, Plan Reading and Construction Techniques. Abingdon: Routledge, pp. 381–382, 387–390. Summarized and paraphrased. Full reference under Supplemental Sources.

Earned Value Analysis of a Construction Contract

Sears et al. point out that a standard monthly cost forecast shows only whether the job will finish over or under budget. It cannot tell whether the variance comes from cost performance, schedule performance or both. The earned value management system (EVMS) separates the two by comparing three cumulative quantities at each update:

• BCWS (budgeted cost of work scheduled): the planned value of the work that the baseline schedule calls for by the update date.
• BCWP (budgeted cost of work performed): the earned value, meaning the budgets of the work actually accomplished.
• ACWP (actual cost of work performed): what the accomplished work actually cost.

From these: cost variance CV = BCWP − ACWP; schedule variance SV = BCWP − BCWS, expressed in dollars; CPI = BCWP/ACWP; SPI = BCWP/BCWS. An index below 1.0 calls for management attention. The estimate at completion is bracketed by EAC₁ = ACWP + (BAC − BCWP)/CPI and EAC₂ = ACWP + (BAC − BCWP)/(CPI × SPI), where BAC is the budget at completion. The to-complete index TCPI = (BAC − BCWP)/(EAC − ACWP) gives the cost efficiency needed on the remaining work.

Source example (highway bridge, day 27): BAC = $390,214; BCWS = $160,622; BCWP = $147,011; ACWP = $157,582. Then CV = −$10,571 and SV = −$13,611; CPI ≈ 0.93 and SPI ≈ 0.92. The forecast range is EAC₁ ≈ $418,273 to EAC₂ ≈ $442,409. In other words, if the cost inefficiency continues, the job finishes about $28,000 over, and continued schedule slippage could raise that to about $52,000.

The authors also describe the time-cost envelope (early-start and late-start cumulative cost curves) as the graphical forerunner of EVMS.

For railroad work: On a unit-price subgrade rehabilitation or slope repair, use the bid items as the budget lines and the construction monitoring quantities (cubic yards of undercut, tons of subballast or riprap, linear feet of drain, piles driven) as the earned value. Because work is confined to track windows, a low SPI often means windows were lost or shortened, not that crews were slow. Keep the window log next to the EV report so the cause is clear. A falling CPI combined with deeper-than-planned undercut is an early quantitative sign of a differing-condition trend that should be raised with the railroad before the next window, not at close-out.

Source: Sears, S. K., Sears, G. A., Clough, R. H., Rounds, J. L., and Segner, R. O., Jr. (2015). Construction Project Management: A Practical Guide to Field Construction Management, 6th ed. Hoboken, NJ: John Wiley & Sons, pp. 298–303. Summarized and paraphrased; figures reproduced for internal reference. Full reference under Supplemental Sources.

Delay Analysis: Gross and Net Extensions, Concurrent Delay and Constructive Acceleration

Haidar reviews how extensions of time are calculated when an excusable event occurs while the contractor is already late through its own fault (culpable delay). Under the gross method, the extension runs from the date the event occurred, which sets a new completion date and removes liquidated damages up to it. Under the net method, the contractor receives only the time needed to absorb the event, added to the original completion date. In his example, a contractor six months late is delayed one further month by a relevant event. Under the net method the completion date moves one month past the original date, and five months of culpable delay remain exposed to liquidated damages. The net approach follows Lord Denning’s reasoning in Amalgamated Building Contractors v Waltham Holy Cross (1952).

On concurrent delay, Haidar cites Henry Boot v Malmaison (2000). If a week of work is lost both to exceptional weather (an excusable event) and to a labor shortage (not excusable), the contractor may still be entitled to a one-week extension. He warns that retrospective CPM analysis of very large networks is unreliable, because every logic link is an assumption and a network can be adjusted to make almost any work critical. Liquidated damages work only if the contract also provides a workable extension mechanism. To recover for constructive acceleration, a contractor must show that the owner failed to grant an adequate extension, insisted on the original date, received notice that the contractor was treating this as acceleration, and that extra cost was actually incurred. The cases cited are English, so the governing contract and jurisdiction control on any U.S. project.

For railroad work: the geotechnical field record is often the evidence in these analyses, so record in daily reports:

• Track time requested, granted, and cancelled or shortened by operations, with times.
• Weather and site conditions that stopped earthwork, compaction testing or drilling.
• Hold points imposed because of monitoring results (inclinometer movement, settlement rate, pore pressure) and the time each was released.
• Contractor resource shortfalls observed (crews, equipment, material) on the same days.

With all four recorded side by side, the railroad and the contractor can separate excusable delay from culpable delay without reconstructing events months later.

Source: Haidar, A. D. (2016). Construction Program Management – Decision Making and Optimization Techniques. Cham, Switzerland: Springer International Publishing, pp. 75–79. Summarized and paraphrased; figures reproduced for internal reference. Full reference under Supplemental Sources.

Risk Monitoring, Control and Contingency Drawdown

Cretu et al. stress that risk management fails most often after the analysis is finished. They invert the usual pyramid: planning and analysis take less effort than carrying out and tracking the responses. Their monitoring and control step includes tracking every risk in a master register, identifying new risks and retiring old ones, adjusting responses that are not working, managing the risk reserve (contingency) as it is spent, and checking that responses are actually carried out. Each risk gets an owner and a task manager who tracks how well the response works and reports on a set schedule. Updates go back to a single master register held by the project manager. The response action plans can be loaded into the project schedule as dated activities so that overdue responses stand out at each status date. On large projects the authors recommend formal risk reviews at the end of concept development (where “project-killer” risks surface), during preliminary design and during final design (construction and contract risks), with interim reviews during construction.

They close with keys to success from NCHRP Report 658. These include: identify risks early and tie contingencies to explicit risks; communicate cost uncertainty as ranges or explicit contingency; assign responsibility for each risk; and monitor threats and opportunities as the way to resolve contingency. They also note that responsibility shared by everyone tends to end up with no one.

For railroad work: Construction monitoring data are the evidence for retiring or escalating geotechnical risks. Link each instrument and test program to the register entries it tests. For example, inclinometer and piezometer readings during staged fill test the “foundation instability” risk, and proof-rolling and density results test the “unsuitable subgrade quantity” risk. Set the threshold that would trigger the planned response. Report contingency drawdown against those same entries at each progress meeting, so the railroad can see that the reserve is being spent on risks that actually occurred or released as they are retired, not absorbed by unrelated changes.

Source: Cretu, O., Stewart, R. and Berends, T. (2011). Risk Management for Design and Construction. Hoboken, NJ: John Wiley & Sons, pp. 149–150, 164–168. Summarized and paraphrased. Full reference under Supplemental Sources.