The excavator sticks precisely to the utility map. The pipe is still over there somewhere.
There was no carelessness in this situation. The operator excavated at the position indicated by the line. That line had been taken from a record, and the record in turn came from a drawing which someone had made many years before and which had never been checked against the trench. At the moment, a team is gathered around a broken main and the schedule has just been missed by weeks.
It is by no means uncommon. In August 2026, the Common Ground Alliance (CGA) estimated the annual cost of underground utility damage in the United States at around $83.2 billion, of which direct repair expenses amount to $4.9 billion; the remainder is due to delays, outages, road closures, and injuries.
Now, allowing for that, many strikes occur before the map has even been opened. CGA's 2025 data continues to indicate that not contacting 811 prior to digging is still among the main root causes. This article is concerned with the other category: the call had been made, the marks had been placed, the map was on screen, yet the position was still incorrect. We looked at the broader issue of the GIS gap responsible for utility strikes in a previous post. In this case, we are looking more closely at the source of the position error, what the term 'located' actually means, and how confidence should be displayed on a map rather than pretending that every line is equally accurate.
Is a mapped utility identical to a located utility?
Not at all. And the difference is where most of the trouble sits.
A line in GIS looks sure of itself
A vector line maintains the same sharp width no matter if it was derived from a surveyed pothole or from a hand sketch made in 1974. The software has no doubts about this. When the reader looks at the line, they assume that it must have been measured.
Mapped versus located
To say that a position has been mapped means that the position was obtained from records—such as drawings, as-builts, old CAD files, or from someone's memory. If it is said to have been located, then someone has gone to the site, detected or exposed the utility, and has linked that position to survey control. These are two entirely different things, and introducing a layer that blends them without making it clear is asking the following excavator to guess.
Where the position error comes from
Records that were never checked in the field
The fact sheet produced by Canada's CSA S250 makes it clear that a great many of the existing map records are inaccurate, out of date, or are in formats that are incompatible with one another. That statement comes from a Canadian standards organisation, but anybody who has looked at a legacy utility layer will be able to see what it means.
Digitizing paper
Paper maps tend to stretch, fold and fade. Research into utility cadastres that were created from scanned paper showed that scanning irregularities, deformation of the map material, and drafting errors caused the digital version to have a varying quality across the map.
Saline, a city in Michigan, illustrates what this kind of situation looks like in real life. In the process of turning decades' worth of paper as-builts — some of which are more than 100 years old — into digital form, it was not at all uncommon to discover that assets were located in the wrong place, for example, a hydrant being on the wrong side of the street. A seasoned maintenance worker picked up a large number of these mistakes by noting the discrepancies that he was aware of from his experience in the field. That is a useful solution, but observe that it relied on one person's memory.
Coordinate transformations and units
It is possible to draw a line correctly yet have it end up in the wrong location following a change to the datum, due to an incorrect projection or because of the wrong units. Here is a simple example. The US survey foot and the international foot differ by two parts per million, which is approximately 0.01 ft per mile. Although this seems insignificant, over state plane coordinates covering hundreds of miles, the difference can total several feet. Since 1 January 2023 the term "foot" in surveying and mapping has officially meant the international foot, and so legacy data that is simply labelled "foot" may inadvertently combine the two. If you are uncertain about what units are actually being used in a particular dataset, our Geo Converter is a sensible option for testing a transformation before it is used in a design.
Changes nobody wrote down
There were relocations, some repairs, certain lines were abandoned, and a new service was put into operation on a Friday afternoon. The map would then show a trench which no longer exists if the record had never been updated.
What do ASCE 38 quality levels really mean?
ASCE 38 is the standard guide used in the United States for gathering and showing existing subsurface utility data; however, as to whether it applies to your project, you should refer to the project specifications and the owner. The concept behind it is simple, namely that each utility should be labelled according to how its position was determined.
D, C, B and A in plain terms
- QL-D is based on records, oral accounts, and one-call markings. There has been no field verification.
- QL-C includes a survey of the visible surface features such as manholes and valve boxes, together with the records.
- The term QL-B refers to the situation where the utility has been detected using geophysical methods and its position has been marked; this procedure is known as designating.
- The fact that the utility was physically exposed and measured is indicated by QL-A.
A summary of ASCE 38-22 states that the survey tie for QL-A is 0.1 ft vertically and 0.2 ft horizontally, and QL-B is also tied to the project survey datum. Make sure to read the standard itself before entering any figures into a contract.
What a level does and doesn't say
The quality level refers to the manner in which the information was obtained, not merely to the accuracy of the survey. This point is easily overlooked. A manhole that has been surveyed represents a precise point, but the pipe between two manholes remains an interpretation.
The same reasoning holds true in the case of QL-A since it is present at the point of exposure. In Australia, the linework between exposure points remains at the level that has been nominated. The same source, a utility-mapping company compiling AS 5488, also gives tolerances of ±300 mm horizontally and ±500 mm vertically for QL-B and ±50 mm for QL-A. Since the standards differ according to region, it is necessary to verify which one is actually being used by your project.
Why does not knowing the exact position affect excavation decisions?
Clearance and planning
A line that is known to an accuracy of a few centimetres and one that has been copied from an old drawing must be treated differently when carrying out the work. In the case of the first, you can set an offset with a certain level of confidence; with the second, you should expose it before the bucket comes near. The rules regarding tolerance zones and hand-digging differ from one jurisdiction to another, so you should verify which ones apply in the place where you are working.
Clash detection
The design teams carry out clash tests on the utility layers and regard the lines as being exact. A QL-D line that passes the test has passed against a guess, and a QL-D line that fails might cause a redesign for a conflict that is not actually there. Both types of error result in financial cost, although the timing is different.
The cost logic
The most well-known figure is based on a study that was commissioned by the Federal Highway Administration. In four states and involving 71 projects, it found that $4.62 was saved for every $1 invested in subsurface utility engineering. A subsequent paper from ASCE, which references the study, states that the cost of the QL-B and QL-A data was less than 0.5 percent of the construction cost, with savings amounting to about 1.9 percent. Although this is outdated data specific to highways, your project might be different. That said, the general trend is difficult to dispute.
How do you check a utility map to make sure it's correct before anyone starts digging?
Record the evidence, not just the line
Include fields for the source, the method of capture, the date on which it was captured, the quality level, the horizontal uncertainty, and the person who verified it. A thesis concerning utility records in the Dubai Municipality suggests that for each asset, the method of capture, its lineage, and the level of confidence should be recorded, as this facilitates selective resurveying. The same study states that positional error can be reduced to about 5 cm by combining different sensing methods. It is not necessary to resurvey everything; one only has to decide which lines should be resurveyed first.
Show it on the map
Represent the quality level by using different types of lines: solid and thick lines for grades B and A, thinner dashed lines for grade D, and include a legend to explain the meaning of each type. It shouldn't be necessary for anyone to open the attribute table in order to realise that a line is an estimate. If you are planning on creating a layout of this kind, then our cartography services deal specifically with the production of maps of this type.
A practical workflow
Well, what exactly do you do? The sequence that we would follow is this:
- Look at the records, gather all the information, record the source together with its date, and identify any conflicts concerning the owners.
- Survey the visible features such as manholes, valve boxes, poles, and markers. This establishes your QL-C and ties the records to actual locations.
- Use geophysics. A qualified utility locating service will detect and mark the lines (QL-B).
- Expose the utility at the key points; for example, at the crossings, in the narrow corridors, and in any other spot where an error would be most damaging (QL-A). This aspect is typically handled by specialists in subsurface utility engineering.
- Make all the work refer to the project control since that is where a topographic survey plays a role. For instance, on the Warsak Wind Farm project, we established the control network by placing concrete monuments and permanent benchmarks and then passed the coordinates on to the engineering consultant and the client. That is exactly what 'tied to project control' means in practice.
- Send it back to the authoritative GIS. Update the geometry, the quality level, the date, and the source. If the new position never makes it to the master layer, then the next team will start from the previous guess.
Final thoughts
A line on a utility map represents a claim rather than a measurement, and there are three things which make that claim usable. Explain how the position was determined, illustrate it on the map so that people won't have to look for it themselves, and check at the points where an error would be most costly before returning the results to the layer.
If you've inherited a utility layer, then my recommendation is straightforward: select the three crossings where an error would cause the greatest harm and check those first. The other parts of the layer can be improved later on, provided it is honest about what it is.