Below is a classic surveying horror story that revolves mostly around the confusion of the Survey Foot (SFT) and the International Foot (IFT) definitions. Dr. Michael Dennis, PS, PE, the project manager for NGS’s 2022 State Plane Coordinate System (SPCS) project originally asked for stories about how SFT in an IFT state caused problems. This story was originally shared anonymously, and most of the details were left out, but because of public interest and so many follow up questions about how such a thing was technically possible we decided to tell more details about what happened. As with many stories of disaster, there isn’t exactly a single culprit, but more over a culture of behavior and sequence of several events that results finally in calamity. Some of them, in hindsight, are obvious, but please consider that at the time, they were probably unknown to the participant as even a source of potential danger.
Just like another great engineering and technical disaster, the sinking of the Titanic, in this story, the SFT is the iceberg, as it were, but looking deeper reveals several other underlying problems that certainly set the stage for the final outcome. Thankfully in this story, there is no loss of life, but there was a permanent financial impact to a high-rise building that exists to this day. This story happened a few decades ago, and because it involved so many firms and because there were settlements and financial loss, most specific details about firm names or even where this project is located are being kept anonymous. We, as an industry and society can learn from our mistakes, but only if we bring them to light and make conscious decisions about them in the future. So let’s dive right in and see where it all began.
The project, a twenty-story high rise building in an anonymous City downtown center only some 12,000 feet from the end of an anonymous international airport runway.
Five companies, one project and a sequence of compounding errors and miscommunications that result in a permanent economic impact to the final product, the loss of an entire floor on a twenty-story building.
First, an out of state company aggregates data from multiple sources. We’ll call them company ‘A’. The aggregated data included the east-west FAA runway approach glide path data, county data with local platting and geographically referenced section corner data, and the building’s architectural concept. Company ‘A’ also contracted with a local firm, Company ‘B’ to perform an ALTA survey of the future building parcel. Some assumptions were made, and some rules were followed, most notably that the FAA mandates their survey data is reported in SFT. The assumption was that it would be obvious that all data was on this system, as it had FAA calculations on it.
This unfortunately has been quite common in states where the IFT is legislated and firms following agency guidelines or from SFT states come in to interact. The underlying geodetic reference implicitly was the FAA survey PAC/SAC control. Again, this was by FAA policy. It was also assumed that the PAC/SAC data would match exactly with the FBN data, as they were both in the NGS IDB.
Company ‘B’s parcel boundary had a misclosure, probably from a scrivener’s error on the south line of the parcel. That will come to play later, but for the moment was unnoticed. The magnitude of the error was 0.15 feet. They also performed the work with terrestrial equipment and the only tie to geodetic control was to section corners that had previously been tied to geodetic control by a County survey. This is principally how Company ‘A’ was able to correlate the ALTA survey (using county section corner positions) to the FAA data.
Next, another local State company continues the project for final design and construction layout. We’ll call them Company ‘C’. Again, assumptions are made, and local rules are followed, most notably State statutes reference the use of IFT for reporting SPCS as does the county. The assumption is that the data provided by Company ‘A’ is all in IFT. For this story we are also going to make the assumption that the meta data (the information about the coordinate system definition and control) was either sparse or simply non-existent form Company ‘A’ to Company ‘C’.
Some of the details of this account have been lost to posterity, but in this era, having high quality survey reports and meta data were very much not the norm. You were lucky just to get a CADD file to make sense of the design, let alone knowing where the coordinate base was for the file. Also, the likelihood of having the geodetic control in the same CADD file as the design was basically nonexistent, as a norm. So, the only data provided by Company ‘A’ to Company ‘C’ are absolute positions of the building corners so that they comply with FAA glide path limits. The data was only given in SPCS northing and easting and a linear unit was not specifically stated other than to simply call it “feet”. Company ‘C’ chose to use the County FBN control as the underlying geodetic for their construction control.
Again, this is by State statute and County policy. It is also reasonable practice that one could argue is a typical standard of care in the industry. Why would a company go through all the hassles of airport access to points that are some 20,000 plus feet away, when there is a County control point within a few thousand feet of the project? In the last step of this story, there was a measured variation in the FBN versus the PAC/SAC control, but the magnitude probably wouldn’t have caused a problem by itself, but it does accumulate and contribute as we will mention later in the conclusion.
Next, a City surveyor who we’ll call Crew ‘D’, performs utility/ROW work adjacent to the project, and uses immediately adjacent cadastral control (road centerline monuments delineating the block of the parcel where the building is located). This was work done in accordance with City and State policies and procedures. The work was not georeferenced to any part of the NSRS and was performed using terrestrial equipment. The ROW work identifies that the building was out of position on the parcel and a potential encroachment into the easement existed. This was further compounded by the fact that only the underground garage was over the building setback line (BSL), making it even more complicated to observe at street level. From street level only a mass of concrete was able to be observed in the bottom of a trench, and the structural question became “how thick is the wall” that can only be observed from inside the structure?
Lastly, a final local survey firm, we’ll call them Company ‘E’ was contracted to forensically evaluate all of the data and make determinations as to what was going on with this project. All the survey control, both County FBN tied control and the FAA PAC/SAC control was tied in a single GPS control survey. The current evidence for the section corners and remaining local block corner monuments were also tied into this survey as were the mapping control points used to topographically scan the above and below ground features of the building construction project, now underway. A 3D map and BIM was created and all the governing spatial limits to the building project were overlaid. These include the ROW and parcel boundaries, the BSL and the FAA glide path.
It became immediately apparent that not only different (and subtlety variant) control was used for all parts of the project, but also different equipment and lastly, different computational parameters including the linear unit SFT v IFT. Also, each phase of the project was separately constrained in some way. Company ‘A’s survey was tied to the FAA PAC/SAC control and reported in SFT, again this was by FAA policy. Company ‘B’ tied their survey to section corners, did not specify linear units and were not using SPCS, again in compliance with ALTA procedures. The County had previously tied the section corners to FBN control and reported the SPCS in IFT, again, following State statutes. Company ‘C’ used FBN control principally to control the absolute position of the building and with a lack of documentation otherwise assumed the SPCS to be reported in IFT. Crew ‘D’ used local monuments to position their work on the street but were unable to detect if they had an absolute error to the overarching georeferenced County cadastral survey in the area.
The analysis also revealed that all of the control had subtle, but measurable variations. Some compounding, some compensating (just to keep all of this interesting). The PAC/SAC data did not match exactly with the FBN data at around the 0.1-foot level of magnitude. The Section corners were not perfectly aligned with the FBN data and it was probable that one or more marks had been rebuilt in a slightly alternate location at the 0.3-foot magnitude. Lastly the parcel corners were obliterated at the time because of construction and it was highly probable that the road centerline monuments were both new and also in slightly different locations compared to previous survey work, again at the 0.3-foot magnitude. This of course was further compounded by misclosure on the ALTA on the south line. End result after all of this is comingled, Crew ‘D’ shows the building to be about a foot or so south of the “correct” location (remember they are measuring to an irregular concrete mass in a trench that represents the outer limits of an underground structure.
Arguably, Company ‘B’ and Crew ‘D’s surveys, being terrestrial in nature, were not impacted by the linear unit variations as much (or at all) as the georeferenced surveys were. We will explain why in a moment. The variation in SFT to IFT is only 2 Parts Per Million (PPM). A terrestrial measurement of a few hundred feet, or even a slope calculation of 12,000 feet is hardly impacted in magnitude, and often is a number that is far less than the error of the measurement in the first place. A 2 PPM error in 12,000 feet is only about a quarter inch and certainly is negligible when compared to the inch size error tolerance in this same distance when using terrestrial equipment. The first two surveys were impacted however because they were using SPCS coordinates with magnitudes in the millions of feet (the distance back to the “calculated origin of the zone” millions of feet away).
The northing in this example was right around one million, so the difference “on the ground” equates to right around two feet in absolute position. In other words, a single geographic position (like the calculated corner of a building in relation to the bounds of the FAA glide slope) represented using a single SPCS zone definition, but two different linear units, results in two different “points on the ground”. Those two points are located north-south of each other by two feet. The building, being east west from the airport, was designed to be just north of the glide slope boundary, thus was teetering right on the line of encroaching into that prism of approaching planes. It also put the building over the BSL lines and encroached it into the utility corridor. This was further compounded by the fact that only the subterranean parking garage encroached, where the above ground “walls” were well back from the roadway.
Where does the blame lie and who made the mistake? The highest culpability, in my opinion, is with Company ‘C’. They should have further verified that the absolute position of the south end of the building stated by Company ‘A’ with SPCS was indeed in IFT as they assumed. Company ‘C’ could have further verified the BSL relation to the walls and caught the error before construction. In reality everyone was partially to blame: Company ‘A’ for lack of metadata, Company ‘B’ for closure exceeding ALTA standards and for not calling out the positional variations in the section corners when compared to the County survey, and I’ll assign some blame too to the City for not having a modernized policy in place for their field crews to at least uncover that the local perpetuation of the road centerline monuments are not in original locations.
Ironically, we, as the final firm, were called in to resolve the boundary and setback problems at ground level, which we, to my recollection were able to do with some effort (variances, rewriting easements, etc.) but we also uncovered the problem that the building was principally 2 feet south of where it was “intended” to be on the first design, relating to the FAA glide slope. The intent also was to have the building just north of (my memory seems to think it was half a foot) and slightly above the flightpath (I think it was 4 – 6 feet above).
In reality it was within the flight path and was solidly a vertical encroachment because it was several feet—not fractions of an inch—as one might initially assume from such a small 2 PPM variation in units. So, we resolved all the “on the street” issues of the project and uncovered the “in the air” problem that no one was initially aware of. There was no revising this as the rooftop fixtures, equipment and architectural elements were “set in stone” so the only other outlet for correction was to eliminate an entire floor and shorten the building by fifteen or so feet to eliminate the multi foot vertical encroachment at the top caused by the horizontal shift, directly attributed to the SFT error.
In brief, this was mostly a horizontal problem from a million feet away and not a vertical calculation problem of a slope that was only a few thousand feet long. The final and long lasting 5% loss of real estate came from the “ice burg” of the SFT in an IFT state.
Recent I attended a companywide safety meeting. Our safety director was reporting on a rash of fender-bender accidents we as a workforce were starting to have. Whereas a fender-bender isn’t much of a major problem by itself, it is however a cultural indicator of potential complacency or misunderstanding of underlying threats. These canaries in the coal mine can be the harbinger of much worse things to come, such as a vehicular accident resulting in a fatality. We as responsible professionals need to take heed of warning signs and plan our future actions to be on a better trajectory. Deprecating the SFT form the National system is a long awaited and necessary action but there are other cautionary tales to be taken away from stories like this one as well. Missing meta data or non-existing survey reports are equally as problematic as using an incorrect definition of a foot. Also non-standard actions of Federal, National and State level agencies can cause conflicts and problems as well.
The real moral of this story is that there were several survey groups, all with a mixture of equipment, local knowledge, basic skill, etc. Ironically it was the least skillful/advanced guys that brought the problems to light in the first place by holding local road monuments and discovering that the building was “off”.
Only after very diligent, comprehensive and somewhat costly analysis by the final firm did the reality of the situation fully come to light. The adage that if you don’t have the time to do it right in the first place, you better have the time (and money) to do it over, unfortunately comes to mind.
As a final thought, the so-called grid-to-ground problem of SPCS (often measured in 100 – 400 PPM magnitudes in the taller western states) has historically caused many times the problems on projects when compared to the 2 PPM incorrect foot. I’m so happy that LDP designs in the forthcoming 2022 SPCS “could” eliminate this, but I’ll bet there are still 20% of surveyors that will insist on scaling a project to get rid of that last 15 ppm of linear distortion on their 660 foot long lines (for perspective, that’s a difference of only an eighth of an inch in the length of that line, but 15 feet of horizontal translation to the so called “ground” coordinate, at the one million foot magnitude).
Let’s just hope that we all can learn the lessons and necessity of documentation. I am thankful that NGS and NIST are finally addressing the SFT and instilling only the IFT for future use by all agencies. I am also thankful for the great work that NGS is doing regarding the 2022 datum and all the data products associated with it. I am hopeful that we as an industry can additionally implement standards to go along with all these wonderful tools we are getting so we can get to the root of our professional function, the commerce and safety of the public we are here to support.
A takeaway that I personally implemented in all my current work is to never simply report the SPCS northing and easting in a report or on a survey. I report both the geographic and Cartesian coordinates for at least two points on the survey (usually the primary control) and I follow that with all the parameters of the coordinate zone, including the linear unit. I’ve been teased occasionally for my over exuberance in detail, but it is horror stories like the one above that put it in most perspective for me.
Brian Fisher, PS, is a surveyor for the Central Arizona Water Conservation District and is in charge of geodetic and structure deformation surveys. He is also the Arizona State Geodetic Coordinator for the National Geodetic Survey.