Experienced civil engineer interviews skip definitions like curing and ask what you decided when a pile test came back odd, a pump failed mid-pour, a neighbour's building started to move or a client wanted another floor. Expect questions on foundations and groundwater, mass concrete, temporary works, design coordination, claims, cost tracking, safety investigations and coaching junior engineers. This page is written for civil engineers with roughly three to ten years on real projects, who run a package, a block or a whole site and sign off other people's work. Every answer below is a first-person story with the situation, what you did, the result and the trade-off you accepted. Swap in your own projects and numbers before you say it.
Search all questions by round, difficulty and level, or save the ones you want to practice.
Today: the approvals that are yours alone, like pour cards, measurement or method statements.
The limit: what goes up the line, and why that's right for now.
Next step: the authority you want next and what you've done to earn it.
“Right now I sign the pour cards for my block on my own, so I'm the last check on reinforcement, embedded items and formwork before concrete goes in. I also certify the monthly measurement for three subcontractors and approve their method statements for routine work. What still goes to my project manager is anything that changes cost beyond my limit or changes the design, and temporary works above a set category go to the temporary works coordinator. I think that split is right, because those calls need someone who sees the whole contract. The line I want to move is commercial. I'd like to run a package's cost and variations myself, so for the last year I've prepared the variation assessments for my manager to check, and most now go through unchanged.”
Claiming you sign off everything yourself, or being unable to say where your authority actually stops.
Find the bottleneck: track each activity in the cycle and see what really waits.
Change the flow: prefabrication, extra formwork sets, crane planning, parallel work.
Protect limits: stripping and loading still follow the agreed criteria and the designer's back-propping plan.
“On a fourteen-storey frame we were at about nine days a floor, and the programme needed seven. I tracked a few cycles hour by hour and found the crane was the real bottleneck: rebar bundles, formwork and column cages all fought for it in the same morning. So we moved column and beam cage fabrication to the ground and lifted finished cages, planned crane time as a daily schedule, and bought a second set of column forms so columns on the next floor could start while the slab was still being fixed. We didn't touch stripping times. Those stayed tied to field-cured cube results and the designer's back-propping plan. We got to seven days by the sixth floor. The trade-off was more yard space and a larger rebar crew early on.”
Claiming a faster cycle that came from striking formwork early or skipping inspections.
The gap: what they were missing and how you noticed.
How you coached: doing it together, then letting them lead with you checking.
Result: how their work changed, and when you stepped back.
“I had a graduate engineer who signed off slab reinforcement quickly and missed things, like wrong cover blocks or missing extra bars at openings. He wasn't careless; he was checking against the general arrangement and didn't know what to look for. So for two weeks we did every inspection together, and I talked through what I was looking at: the bar schedule, the laps against the drawing notes, the chairs, the cover, and anything around openings. Then he led the inspections and I followed a few minutes behind and checked, and we compared notes. I also had him write his own checklist from what he'd learned rather than handing him mine. By the second month his sign-offs matched mine, and I trusted him with a whole block's slabs. He now trains the new graduates the same way.”
Saying you simply redid the junior's inspections yourself, or reported them without trying to teach them.
The idea: what you changed and why the original way was costly.
Approval: who had to agree, including the designer of record.
Trade-off: what the change cost elsewhere, and the net result.
“On a housing project the stair flights were cast in place on every floor, and they sat on the critical path because the crew had to form, fix and pour them after each slab. I proposed precast flights made in our yard. I took it to the structural designer first, because the landings and connections had to be redetailed, and to the client with a simple time comparison. Once approved, we saved roughly two days per floor on the frame and got safe stair access much earlier. What we gave up was flexibility. Every flight now depended on the tower crane, so on windy days they waited, and one batch was cast with the wrong riser height because a drawing revision didn't reach the yard. After that I made the yard sign the same drawing register as the site.”
A saving that came from reducing cover, grade or reinforcement without the designer's written approval.
Baseline: what the site investigation and contract said the ground would be.
Evidence: records made at the time, such as photos, logs, measurements and notices.
Judgement: what an experienced contractor could reasonably have expected, and the fair time and cost.
“I was the client-side engineer when a contractor hit a band of boulders while excavating for a basement and claimed for extra time and cost. My first step was the site investigation report. The boreholes near that corner showed dense gravel, but nothing about boulders, so something unexpected was plausible. What made it fair was the records. Because they'd notified us early, I'd had my engineer photograph and log the excavation daily, measure the volume of boulders removed, and note the plant used. I accepted the extra excavation and breaking for that zone and a matching extension of time. I rejected the part of the claim for the whole site, because the records showed normal digging elsewhere. It settled quickly because both sides were arguing from the same daily log.”
Rejecting or accepting a claim on instinct or pressure, with no reference to the site data, the records or the contract.
Three numbers: planned value, earned value for work actually done, and actual cost.
What they show: earned against actual gives cost performance, earned against planned gives schedule performance.
Use it: measure progress honestly, look at trends by package, and act early.
“Money spent on its own tells you nothing, because you can be under budget simply because you're behind. I use earned value. For each package I have the planned value, what we should have done by now at budget rates, and the earned value, what we've actually built at those same rates, and I compare both with the actual cost. If earned is below actual, we're paying more than planned for the work done. If earned is below planned, we're behind. On one project the total looked fine, but splitting it by package showed the basement waterproofing was badly over on cost and hiding behind savings on the frame. We found the crew was redoing sections of membrane because it was being damaged during backfill, and fixing the backfill method brought it back in line.”
Saying a project is on budget because spending is below the plan, without asking how much work has actually been done.
Heat: mix with lower heat, such as cement partly replaced by fly ash or slag, and cooler concrete at placing.
Temperature control: sensors at core and surface, insulate the surface, watch the difference, not just the peak.
Continuity: pour sequence, layer thickness, backup pump and plant so a live face never goes cold.
“On my last raft, about two metres thick, I treated heat as the main risk. We agreed a mix with a good share of slag to cut the heat, and we poured overnight with chilled water in the batching to keep the placing temperature down. I had temperature sensors cast in at the core, mid-depth and near the top, and the spec set a limit on the difference between core and surface. The trick is you don't cool the top with cold water, you insulate it with blankets so the surface stays warm and the gap stays small. For cold joints, we poured in layers along the length with a set time between visits to each area, two plants on order and a standby pump rigged. The difference stayed inside the limit and we saw no thermal cracks.”
Planning to cure a mass pour with cold water on the surface, which widens the temperature gap and causes the cracking you're trying to prevent.
Buy time: keep the face workable, turn back or hold trucks, check the mix for retarder.
Decide: if the face will set before the pump arrives, form a proper construction joint.
Place and treat it: a location the designer accepts, a straight stop-end, then prepare the joint before restarting.
“First I'd stop more trucks leaving the plant and ask what retarder is in the mix, because that tells me how long the face stays workable. Two hours in warm weather is too long to count on, so while we wait for the pump, I'd plan for a construction joint. I'd call the structural designer, because the joint should sit where shear is low, usually in the middle part of the span and square to the main bars, not wherever the concrete happens to stop. Then the crew forms a straight stop-end there and compacts the concrete properly up to it. Trucks already on site go into the slab to reach the stop-end or get rejected. Before the next pour we remove the weak surface layer, roughen and clean the joint and wet it. I'd record the joint position on the as-built drawings.”
Letting the concrete stop wherever it runs out, leaving a sloping cold joint in a random position with no record.
Trace: map where water appears and when, and check the usual weak points like joints, tie holes and pipe entries.
Fix: a repair that suits the cause, such as injection for a live leak at a joint.
Prevent: change the detail or the inspection so the next job doesn't repeat it.
“Six months after handover, the client reported water in one corner of a basement car park after heavy rain. I mapped every damp patch and found they lined up with a horizontal construction joint in the retaining wall and two formwork tie holes. We opened up one spot and found the waterstop at that joint had been displaced during the pour, so there was a gap. The repair was resin injection along the joint and properly sealed tie holes, and the leaks stopped. The bigger fix was on our side. On the next basement, I added a hold point where the engineer checks every waterstop is fixed, continuous and joined correctly before we close the formwork, and I specified sealed tie systems. It cost a bit more time on each lift, but that basement stayed dry.”
Painting a coating over the damp patches without finding where the water actually gets in.
Stop: hold the cap and anything that would bury the pile.
Evidence: pour records, further testing such as coring or cross-hole tests, or exposing the shaft.
Decision: the designer decides whether to accept, add a pile or redesign the cap, and you record it.
“On a bored pile job, the low-strain integrity test showed a reflection about eight metres down on one pile, which can mean a neck or a soil inclusion. I put that cap on hold straight away and pulled the pour record. It showed the concrete level had dropped while the casing was being lifted, which fitted a neck. That pile had tubes for cross-hole testing, so we ran it, and it confirmed a poor zone. I sent everything to the foundation designer rather than guessing the capacity myself. They chose to add a pile next to it and redesign the cap to spread the load across both. It cost about a week on that grid line. The lesson I took was about casing withdrawal, so I made the pile supervisor record the concrete level against casing depth on every pile after that.”
Judging the pile acceptable yourself, or letting the cap be cast to save time before the designer has decided.
The risk: lowering the water table raises effective stress in soft soils, which can consolidate and settle nearby foundations.
Control: a cut-off such as sheet piles or secant piles, and pumping only inside it, or recharge outside.
Watch: a condition survey first, then piezometers, settlement points and trigger levels.
“The main worry is settlement next door. When you pump groundwater down, the soil below the neighbours carries more of the load itself, and soft clay squeezes and settles. Older buildings on shallow footings don't tolerate much of that. On a job like this I pushed for a secant pile wall toed into a low-permeability layer, so we pumped inside the box and barely touched the water outside. Before we started, we did a photo and crack survey of every neighbouring building with the owners. Then we put piezometers outside the wall and settlement points on the neighbours, with alert and action levels agreed with the designer. Outside water levels barely moved, settlement stayed well under the alert level, and the survey protected us when one owner later blamed us for an old crack.”
Treating dewatering as just a pumping problem, with no thought for what it does to the ground under the neighbours.
Verify: re-read the points, check it's not a survey error, look at the trend.
Act on the plan: tell the designer and your manager tonight, follow the agreed alert response.
Protect: increase reading frequency, pause the riskiest activity, prepare the contingency.
“First I'd make sure it's real. I'd get the surveyor to re-read those points and a couple of nearby ones, and check the trend over the last few days, because one odd reading can be a disturbed target. If it holds, I don't wait until Monday. I'd call my project manager and the temporary works designer that evening and follow the response we agreed in the monitoring plan. Usually that means more frequent readings, and stopping the activity most likely to be causing it, say the next dig stage near that wall or a pump that's drawing down more than planned. I'd walk the neighbouring building for new cracks and let the owner know we're watching it. And I'd make sure the contingency, like extra props or backfilling that bay, is ready to go if readings keep moving toward the action level.”
Logging the reading and carrying on until the next weekday meeting.
Show it: a large-scale sketch or mock-up of every bar at the joint, including laps and stirrups.
Options to the designer: couplers instead of laps, laps moved out of the joint, a change of bar size, a better-flowing mix.
Confirm: revised detail issued, trial fixed, checked before the pour.
“On a podium level, the transfer beams met the columns with column bars, two layers of beam bars both ways and laps all at the joint. On paper it worked, but when I drew it full size there was barely room for a vibrator. I built a quick mock-up in the yard and invited the structural designer to look. We agreed three changes: mechanical couplers for the column bars so the laps left the joint, staggered beam laps moved away from the column face, and a mix with smaller aggregate and higher workability for those pours. The designer issued a revised detail, and I had the first joint fixed and inspected before anyone repeated it. It added coupler cost, but the joints came out dense and we avoided the honeycombing I'd seen on a similar job before.”
Allowing bars to be cut, cranked or left out on site without a revised detail from the designer.
Stop and assess: stop more cutting, find out what was hit, such as bars cut and position in the member.
Designer's call: send the facts to the structural designer and carry out the repair they specify.
Prevent: sleeves before the pour, a permit for any drilling into structure.
“On an office fit-out, the plumbing crew core-drilled a large hole through a main beam near a support to fit a drain line. I found it on a walk-round. I stopped all coring on that floor and got the beam scanned and the hole measured. Two bottom bars had been cut, right where shear is high. I sent photos, position and bar details to the structural designer the same day. They specified a bolted steel plate to strengthen it, and a different route for the pipe. The bigger change was the process. After that, no one drilled into concrete without a signed permit from the site engineer, and for the remaining floors we held a services coordination check before each pour so sleeves were cast in, not drilled later.”
Filling the hole with mortar and moving on without telling the structural designer.
Make it visible: a query log with the date each answer is needed by, tied to the programme.
Make it easy: send a proposed solution with the question.
Escalate with facts: show which activities are at risk, through the proper route.
“On a hospital extension, our queries to the structural consultant were taking two or three weeks, and some stopped the crew. I changed how we asked. Every query went in a shared log with a needed-by date taken from the programme, so an answer due in three days was clearly marked. I also started sending a proposed answer with each query, like a sketch of how we'd detail a clashing corbel, so they could just confirm or correct it. I set up a short weekly call to go through open items instead of more emails. When three items still got close to holding up the frame, I escalated through the client's project manager with the log showing the programme impact. Response times dropped to a few days, and we never built anything on a guess.”
Either stopping and waiting with no follow-up, or building on your own assumption without written confirmation.
Know the slab: original drawings, and a scan for bars, services and any post-tensioned tendons.
Design: an engineer checks the slab around the opening and designs trimming or a steel frame.
Method: propping before the cut, a controlled cutting method, and a check of the edges afterwards.
“First I want to know what the slab really is. I'd look for the original drawings, but I'd never trust them alone, so we scan the area for bars, conduits and, most importantly, post-tensioned tendons. Cutting a stressed tendon is dangerous and can take load out of a whole bay, so if the slab is post-tensioned it becomes a specialist job. Next, an engineer checks the slab with the hole in it, because cutting bars changes how it spans, and designs the trimming, often a steel frame or beams around the opening. On site we prop the slab below before cutting, cut with a wall saw or stitch drilling rather than breaking it out, and check the exposed bars and edges once it's open. On my last job the scan showed a drainage pipe inside the slab exactly where the architect wanted the stair.”
Marking out the opening from an old drawing and starting to cut without a scan or an engineer's check.
Survey first: visual inspection and crack and spalling mapping to decide where to test.
Strength: rebound hammer and ultrasonic tests as indicators, cores as the reliable measure.
Durability: cover depth, carbonation depth, chloride content and half-cell readings for corrosion risk.
“I start with a proper visual survey, mapping cracks, spalls and rust stains, because that tells me where to test. For strength, the rebound hammer only reads surface hardness, so I use it to compare areas, not as a real number. Ultrasonic pulse velocity is good for spotting voids and poor zones. The test I trust for strength is cores, and I place them using the other results. With rust stains, durability matters more. I'd ask for cover measurements, carbonation depth using the indicator spray on a fresh break, chloride tests, and half-cell readings to show where corrosion is likely. On one building the cores were fine, but carbonation had reached the bars on a sheltered side where the cover was thin. The fix was concrete repair and an anti-carbonation coating, not strengthening, which saved the client a lot of work.”
Reporting rebound hammer readings as the concrete's actual strength, with no cores to back them up.
No quick answer: it depends on the original design, not on how strong the building looks.
What gets checked: design loads, columns and foundations, lateral stability, and the as-built condition.
Route: a structural engineer's assessment and the planning and code approvals.
“I'd tell them I can't give a yes on the spot, however solid the building looks. The frame and foundations were designed for a certain number of floors, and an extra one adds load all the way down. So I'd want the original design and drawings, or if those are missing, a survey with cores and scans to find the real sizes, bars and strength. A structural engineer then checks the columns, the foundations against the soil data, and the effect on wind and earthquake loads, which can be the deciding factor. There's also planning permission and the building code for the new height. Sometimes the answer is yes with a lightweight steel floor, sometimes yes after strengthening, and sometimes no. I'd give the client that route and a rough timeline, not a guess.”
Saying yes because the building looks strong or the columns look big enough.
What happened: the set-up, what was wrong, and how it was found.
Immediate action: stop loading, make safe, involve the temporary works designer.
System change: checks and a permit to load so it can't recur.
“On a transfer slab, I did my own walk under the falsework the evening before the pour. Along one edge the props were standing on backfill that had been rained on all week, and a couple of base plates had already sunk slightly. There were also a few missing braces near a service opening. I stopped the pour booking there and then, which wasn't popular with sixty cubic metres ordered. We put in timber sole plates on compacted ground, replaced the braces, and the temporary works designer checked it before we loaded it two days later. Afterwards I brought in a written permit to load: nobody pours on falsework until the designer's checklist is signed by a named person who has walked it, including the ground under it. That permit caught two more problems on that job.”
Treating falsework as the carpenter's business and signing the pour card without looking at what holds it up.
The work: what the method was for and why it mattered.
The flaw: the technical gap, explained simply.
Outcome: what was changed and how you checked it on the day.
“A formwork subcontractor sent a method statement for tall shear walls, about four metres in one pour, using self-compacting concrete. Their formwork was designed for a normal mix placed at a steady rate, so it assumed the pressure would ease off as the lower concrete stiffened. Self-compacting concrete stays fluid, so you should assume full liquid pressure over the whole height, and filling fast gives it no time to stiffen at all. I sent it back asking for the formwork check to use the pressure for that mix and pour rate, and for a limit on how fast they could fill it. They came back with extra ties and a maximum rise per hour. On the day, I had an engineer at the wall timing the rise. The walls came out straight, and we had no bulging or blown ties.”
Approving a method statement because the format was complete, without checking the engineering inside it.
Contain: tag and isolate the rest of the batch, stop fixing from it.
Trace: use delivery notes, heat numbers and the mill certificate to find where any bars went.
Confirm and decide: retest as the standard allows, inform the supplier and designer, replace what's needed.
“First I stop anyone using bars from that delivery and have the stack tagged and moved aside. Then I trace it. Delivery notes and the heat numbers on the tags tell me which batch it was, and our issue records should show where those bars went. If any are in a slab that hasn't been poured, it's simple: we pull them out and replace them. If any are already cast, I tell the structural designer straight away with the test figures, because they decide whether the member is still fine or needs checking or strengthening. In parallel, I compare the mill certificate against our results, ask the supplier for an explanation, and arrange the retests the standard allows, usually on more samples. I wouldn't release the batch on a single passing retest, and I'd increase testing on that supplier's next deliveries.”
Assuming the lab got it wrong and carrying on fixing from the same batch.
Risk-based: hold points where work gets hidden or is expensive to fix.
Clear ownership: who inspects, what they check, what record they sign.
Feedback: track what inspections find and adjust the plan.
“On a mid-rise residential job I wrote the inspection and test plan for the structure. I put hard hold points where work gets buried and is costly to put right: foundation formation before blinding, reinforcement and embedded items before every pour, waterstops before wall formwork closes, and waterproofing before backfill. Everything else, like formwork alignment, became a witness point or a routine check. Each hold point had a short checklist, a named inspector and a signed record, and no pour happened without a signed card. I also logged what the inspections found. After a couple of months that log showed most rejections were cover blocks and missing extra bars at openings, so we spent a toolbox talk on exactly those. Rework on slabs dropped noticeably, and handover snag lists for structural items were much shorter.”
An inspection plan that treats every activity the same, or one nobody on site actually signs against.
Facts: what happened, gathered quickly from people and the scene.
Root cause: why it happened, past the first obvious answer.
Fix: changes to planning, equipment or supervision, and whether they held.
“A bundle of formwork panels slipped from a crane lift and landed on a slab two floors below a working area. Nobody was hurt, but it could easily have been fatal. I led the investigation. The first answer was that the rigger slung it badly, which was true, but not enough. When we asked why, we found the panels were being lifted in loose bundles because the proper stillages were at another site, the rigger was new and hadn't been assessed on that type of load, and there was no exclusion zone under the lift route because the plan had changed that morning. We got stillages back, made lifting of loose materials a supervised lift, and required the lift route to be barricaded before any crane lift. I checked all three a month later, and they'd stuck.”
Ending the investigation at 'worker error' with retraining as the only action.
ClapAssist is an AI interview assistant for Mac and Windows. It listens to the interview on your computer and shows you what to say, in short lines you can read while you talk. Your live interview audio and screen are never stored. Your resume and notes are saved to your account so the app fills them in on any computer. It stays out of screen share on every plan, including Free; only you can see it.