LLD Quality Control Checklist: 12 Fiber Design Errors to Catch Before Construction

Every LLD package that leaves our production desk has to pass a second-engineer review before it goes anywhere near a construction crew. Not a quick look-over — a structured review by someone who didn't touch the design. That independence matters more than people realize. The designer reviews what they intended to draw. The QC engineer reviews what's actually there.

Over the years, running this process across hundreds of fiber projects, we've catalogued the errors that show up repeatedly. Twelve of them appear often enough — and cause enough downstream damage — that they've become fixed checkpoints in our low level design checklist fiber project process. Some are obvious in hindsight. Others are subtle in ways that will wreck a construction schedule if they slip through.

Here's the full list, in order of severity.

What a Fiber LLD QC Review Actually Looks Like

Before getting into individual errors, it helps to understand the structure. Our fiber design QC process isn't a checklist someone runs through in 20 minutes. For a mid-size project — say, 80 route miles with 14 FDHs and a mix of aerial and underground segments — a thorough LLD design review process takes 20 to 35 hours. You're cross-referencing the plan sheets against the splice matrix, the splice matrix against the fiber assignments, the fiber assignments against the BOM, and the BOM against what the HLD actually specified.

The key structural rule: the reviewer cannot be the designer. Period. That's not a procedural nicety — it's the difference between catching errors and rationalizing them.

With that framing, here are the 12 errors we check for every time.

Error 1: Fiber Count Mismatch Between HLD and LLD

The HLD says 144F feeder. The LLD shows 288F on the plan sheets — or worse, 96F. This sounds like the kind of mistake that couldn't possibly make it through production. It does, regularly.

How it happens: the HLD was revised after the LLD work started, and the designer pulled fiber counts from an earlier version of the network topology. Or the LLD was built from a template that had different count assumptions baked in. Either way, by the time this gets caught in the field, cable has been ordered and reels are on-site.

The downstream cost is severe. Reordering a 144F armored ADSS cable for a 6.2-mile feeder run on a 14-week lead time — while a construction crew sits idle — runs the project roughly $31,000 to $47,000 in direct reorder and delay costs, depending on region and contractor day rates. The fix is straightforward during QC: pull both documents side by side and verify every feeder and distribution segment count against the HLD table. Takes maybe 40 minutes. Saves weeks.

QC check: Cross-reference every cable segment in the LLD against the HLD fiber count table. Flag any discrepancy, including differences between plan sheets and the cable assignment schedule within the LLD itself.

Error 2: Missing Spans in the Cable Assignment

Cable assignment schedules — the table that says which cable covers which spans — occasionally have gaps. A segment of 3 to 7 poles where no cable is assigned, because the designer missed it during the build or a route revision removed it from the count.

Construction crews don't always catch this during installation planning. They assume the assignment is complete. The gap surfaces when they're in the field, somewhere on a rural road outside of cell coverage, and the cable assignment simply doesn't continue past a certain pole number. Now they're calling the office, the project manager is calling the engineer, and everyone spends a half-day figuring out what was supposed to go there.

On a project in eastern Tennessee, we caught a 4-pole gap in a cable assignment schedule during QC that would have left 1,240 feet of strand unassigned on a 96F distribution run. The construction package review took the reviewer about 25 minutes to walk the full assignment table. That gap would have been a field RFI at minimum, a re-design and re-pull at worst.

QC check: Walk the cable assignment schedule sequentially from the feeder terminal through every distribution leg. Verify that every span has an assigned cable segment with no breaks. Cross-check against the pole/structure count on the plan sheets.

Error 3: Wrong Attachment Heights

Desktop estimates versus field-measured heights. This one shows up constantly on projects where the field survey data didn't get properly integrated into the LLD, or where the engineer used GIS attachment height assumptions that haven't been ground-truthed.

It matters most at road crossings. NESC Table 232-1 requires a minimum of 18.5 feet of vertical clearance over roadways for communication conductors — and that clearance has to be calculated at the worst-case sag condition, typically full ice load or high-temperature sag depending on region. If your attachment height is estimated rather than measured, your clearance calculation is wrong. Full stop.

We caught a crossing on a rural highway in southern Georgia where the LLD showed a 26-foot attachment height based on GIS pole data. Field survey had the actual attachment point at 22.7 feet, with a sag calculation that put minimum clearance at 17.1 feet — 1.4 feet short of NESC minimum. That's a redesign, a permit revision, and potentially a pole replacement depending on what make-ready options are available. Caught at QC, it's a field survey correction and drawing update. Caught during DOT permit review — or worse, during construction — it's a project stoppage.

QC check: Verify that attachment heights at all road crossings, railroad crossings, and power line crossings reflect field-measured values, not GIS estimates. Cross-check survey data against plan sheet annotations.

Error 4: Splice Case Fiber Assignment Errors

Buffer tube assignments in the splice matrix that don't match what the plan sheets show. Wrong tubes going to wrong distribution legs. This is the error that, if it reaches activation testing, can cost $40,000 or more per FDH in rework — re-opening closures, re-splicing, re-testing every affected span.

The splice matrix is supposed to be the authoritative document for what connects to what inside every closure. In practice, it often gets built early and then not updated when the fiber routing changes mid-design. By the time the LLD is packaged for release, the matrix may reflect a version of the network that no longer exists.

Our splice point placement guide goes deep on how splice locations affect closure sizing and tube management — that's the upstream design work. QC is where you verify the assignments are actually correct. For every splice case, the reviewer checks: does buffer tube 1 on the incoming cable connect to the distribution leg shown on the plan sheet? Does the color-code sequence match the fiber map? Are any tubes listed in the matrix that don't appear in the plan sheets?

QC check: For each splice closure, cross-reference the splice matrix buffer tube assignments against the plan sheet fiber routing diagram. Verify color-code sequences match industry standards and that no tube is assigned to multiple legs or unassigned entirely.

Error 5: Reel Length vs. Span Mismatch

Cable reels that don't cover the assigned spans, forcing unplanned mid-span splices. This one is a construction cost error and an optical performance issue at the same time.

The reel length planning has to account for the actual span length, sag allowance, slack loops at both terminal points, and a 3 to 5 percent buffer for pull tension and routing deviations. When the BOM calls for 5,000-foot reels on a segment that's actually 5,340 feet of route with 80-foot slack loops at each end, you've got a mid-span splice at best, a complete re-pull at worst. A good reel length planning methodology bakes all of this in before the BOM is final — but QC still has to verify the numbers.

We've seen this error most often on projects where route revisions happened after initial reel sizing. The route gets extended by 400 feet to avoid a utility conflict, but the reel lengths in the BOM never get updated. The plan sheets show the new route. The BOM still shows the old reel count.

QC check: For every cable segment, verify that the specified reel length(s) cover span distance plus slack loop allowances plus a minimum 5 percent buffer. Flag any segment where the math doesn't work. Cross-check against the cable assignment schedule for consistency.

Error 6: Missing or Incorrect Conduit Specifications

Wrong diameter. Wrong material. Missing innerduct specs for underground segments. These errors show up most often on projects where the aerial and underground design work was split across multiple engineers without a clear handoff protocol.

A 1.25-inch HDPE conduit specified where the plan requires 2-inch for a 144F cable — that's a contractor field change that costs $8 to $14 per linear foot to re-bore and costs the contractor time they'll bill back. Missing innerduct specifications on a shared conduit with multiple fiber bundles means the installation crew has to make judgment calls in the field about duct assignments, which tend to be wrong in ways that only matter two years later when the first repair is needed.

Conduit specifications should include: outer diameter, wall schedule (SDR-11, SDR-13.5, or Schedule 40), material (HDPE, PVC, HDPE with integrated innerduct), innerduct configuration if applicable, sweep radius at bends, and pull-section lengths. If any of those are missing from the underground plan sheets, it's a finding.

QC check: For every underground segment, verify that conduit outer diameter, material, wall schedule, and innerduct configuration are explicitly specified. Cross-check against the construction package deliverables standard and verify consistency with the cable specifications in the BOM.

Error 7: FDH Port Assignment Conflicts

Two distribution legs assigned to the same port. Or ports assigned that don't exist on the specified cabinet model. Both happen more than they should.

The dual-assignment error typically happens when a designer copies a port assignment block from one FDH to another and doesn't update the port numbers fully. The FDH plan sheets look fine individually — the error only surfaces when you compare multiple sheets simultaneously or against the port assignment register. By the time construction installs the cabinet and a technician is trying to connect the second leg to a port that's already occupied, you're looking at a field RFI and a design revision under time pressure.

The non-existent port error is even more basic. The LLD specifies ports 25 through 36 on an FDH configured with 24-port capacity. The designer was working from a different cabinet spec than what was actually procured. The fix is easy at QC. In the field, it's a procurement and installation delay.

QC check: Pull the actual spec sheet for each FDH model specified in the BOM. Verify that every port assignment in the LLD falls within the actual port count. Check for duplicate port assignments across all distribution legs at each FDH location.

Error 8: Permit Drawing Format Errors

Wrong scale. Missing title block information. Not matching the jurisdiction's requirements. Permit drawing errors are a QC failure that costs time, not materials — but time is the resource that kills broadband deployment schedules.

Different jurisdictions have genuinely different requirements. A state DOT may require 1:50 scale for road crossing details, a city public works department may specify a different title block format, and a railroad right-of-way permit may require a specific sheet size and stamping format. Getting this wrong means a rejection and a 3 to 6 week re-submission cycle — and many jurisdictions won't expedite re-submissions even when the original rejection was minor.

On a highway crossing permit in western Virginia, we caught a set of plan sheets with the wrong scale notation — the drawings were correct, the title block called out the wrong scale value — during QC, two days before the submission deadline. A 20-minute fix. Had it gone out, we'd have faced a rejection and a lost review cycle at the DOT.

QC check: For every permit submission package, verify scale accuracy, title block completeness (project number, sheet number, revision date, preparer name, jurisdiction-specific fields), and match against the specific requirements of each submitting authority. Don't assume your firm's standard template matches the jurisdiction's requirements.

Error 9: BOM Quantity Errors

Materials don't match what the plan sheets show. This is the single most common class of error we find — and almost always, it's because the design changed after the BOM was generated, and the BOM wasn't updated to match.

Route revisions add poles, crossings, or underground segments that don't make it into the material quantities. A splice case gets added to the plan sheets but not to the BOM. The conduit run changes from 1,200 feet to 1,850 feet but the BOM still shows the original length. Our fiber construction BOM guide covers how to structure a BOM to minimize this drift — but structure alone doesn't prevent it when the design process isn't disciplined about BOM updates at every revision.

The financial exposure here is asymmetric. Under-BOM materials means procurement delays, which can idle a crew at $18,000 to $35,000 per day on a large project. Over-BOM materials means excess inventory that someone has to return or absorb.

QC check: Do a line-by-line reconciliation of BOM quantities against plan sheet counts. Splice cases, FDH counts, conduit footage, pole hardware quantities, and cable reel counts all need to match. Flag any discrepancy greater than 2 percent as a finding requiring resolution before release.

Error 10: Missing Slack Loop Specifications

No slack allocated at splice points, FDHs, or mid-span pull boxes. This one's quiet during construction and loud during the first repair.

Slack loops — the stored cable coiled at splice enclosures, FDHs, and pull vaults — exist so that future re-entry doesn't require a new cable pull. Standard practice is 50 to 75 feet of aerial slack per closure, 30 to 60 feet of underground slack per vault end, and 15 to 25 feet at each FDH entry. When these aren't specified in the LLD, the construction crew makes judgment calls. Sometimes they get it right. Sometimes you discover 18 months later that a crew installed a buried closure with 8 feet of slack per side, which is just barely enough to open the closure but not enough to re-splice a broken fiber without pulling a new segment.

Slack specification is also directly connected to the reel length calculation — you can't plan accurate reel sizes without accounting for the slack. The two errors often appear together.

QC check: Verify that slack loop quantities and storage method (snowshoe, figure-8 coil, slack storage cabinet) are explicitly specified at every splice closure, FDH entry, and pull vault. Cross-check against reel length calculations to confirm the slack is accounted for in the material quantities.

Error 11: NESC Clearance Violations at Crossings

Aerial spans crossing roads, railroads, or power lines without proper clearance calculations — or with calculations based on incorrect attachment height data. This is a safety issue, a permit issue, and a liability issue simultaneously.

NESC Section 23 clearance requirements are non-negotiable. For a communication conductor crossing a major roadway, you need 18.5 feet minimum clearance under full ice load conditions. Railroad crossings have their own requirements, and they vary by railroad owner — Class I railroads typically require 23 feet minimum over the rail, sometimes more depending on electrification status. Power line crossings require separation calculations based on the voltage class of the supply conductor.

Our work on NESC compliance covers the structural pole loading side of this. But NESC clearance violations at crossings are distinct — they're about vertical separation, not pole structural capacity. A span can be structurally adequate and still violate clearance requirements if sag at the worst-case temperature/load condition drops the conductor below the required height.

QC check: For every road crossing, railroad crossing, and power line crossing, verify that the LLD includes: field-measured attachment heights, sag calculation at the governing load case (per NESC Table 250-1 or applicable state standard), resulting minimum clearance, and comparison against the applicable NESC or railroad-owner requirement. Any crossing where clearance isn't explicitly calculated is a finding.

Error 12: As-Built Annotation Conflicts

The LLD shows one thing, field photos show another. Design wasn't updated after field conditions changed.

This happens most on projects where field survey is iterative — multiple field visits, some of which capture changes to the route that don't fully propagate back into the design documents. It also happens when construction is partially underway and the design team is still updating the LLD for remaining segments — a common situation on large BEAD-funded builds where different segments are progressing simultaneously.

A pole gets replaced during make-ready and the replacement is a different height than the original. The fiber route shifts 200 feet east to avoid a newly discovered underground utility. A splice point gets moved to a different pole because the originally specified pole failed the loading analysis. If any of these changes don't get reflected in the LLD before the construction package is finalized, the as-built documentation effort becomes a reconciliation nightmare — and the as-built itself may be wrong, which compounds into problems for any future work on the network.

These annotation conflicts are especially important to catch when the LLD will serve as the baseline for GIS as-built records. See our notes on the relationship between LLD accuracy and construction package deliverables for how this affects long-term network documentation quality.

QC check: Cross-reference the LLD against the most recent field survey photos and any make-ready or routing change notifications. For any flagged field condition, verify the design reflects the actual field state, not the originally planned state.

QC Summary Checklist Table

Here's the 12-point checklist in consolidated form — formatted for use as a sign-off document on each design package review.

Every finding gets documented with a reference to the specific sheet, schedule, or segment where the error was found. No verbal corrections. No "we'll fix it in the field." If it's a finding, it gets a corrective action in writing before the package is released.

Running this 12-point OSP LLD construction drawing checklist on every package adds hours to the design production timeline. But those hours are nothing compared to what any single critical-severity error costs once construction is underway. On a project in northern Alabama with a 107-mile build, catching a fiber count mismatch in QC saved an estimated 11 weeks of procurement delay and somewhere around $290,000 in contractor idle time. That's not an outlier — that's a typical outcome when a package gets released without a proper review.

If you're working on a fiber deployment and want a second set of eyes on your LLD package — or you need a firm that runs this process as a standard part of every construction package release — our fiber design services team handles exactly this work. Reach out at info@draftech.com and we can talk through what a QC engagement looks like for your project. We've run these reviews on packages from other firms, on mid-construction corrections, and on as-built reconciliations. We know what to look for.