Case Study 2: Custom Design of a Protection System for a Renewed High-Pressure Main

CEI was the primary design driver for a new 12” High-pressure system in 2016. The evaluation and design of this system experienced several complications with considerations for configuration of the system, selecting an appropriate Protection system, AC mitigation, DC interference inspection, and the addition of tertiary components added to the protection system at the request of our client.

The result of this design was realized in 2017 when the ICCP Rectifier was energized. The construction cost was estimated to be half that required for the several galvanic systems that would have been required otherwise. Additionally, the current required to protect the line fell within 4% of the estimated current, with 180 mA found to sufficiently protect the line. Additionally, the design resulted in effective mitigation of AC potentials in an inductance corridor and the bonus inclusion of some regulator station piping in our protection strategy.

Introduction:

In 2016 CEI began redesigning one of our clients CP systems as the direct result of our investigation efforts of an entirely separate line. At the time CEI was actively engaged in inspecting a system which was observed to have difficulty in achieving protection. The line was a part of an aging series of line sections that had undergone replacement on portions of the pipeline the previous years. In our efforts to evaluate the line we had determined that an abnormally high AC current existed on the line which had gone unnoticed as the majority of which was draining through portions of the line with poor coating and the impressed current transformer-rectifier, respectively.

As we were providing our observations and recommendations to the operator of this line, it was made known that the corrosion department responsible for the adjacent section upstream was also experiencing some unique AC problems. We were then asked to inspect the adjacent line to provide our insight as to the source and solution to the AC phenomena observed. As we conducted our initial records inspection of the line we realized that the adjacent system had also undergone a substantial replacement of approximately 45% of the existing line. The line was identified as a new 12” .375 wall line with Fusion Bond Epoxy (FBE) type coating. New test points were installed, including new test stations that replaced test points and corresponding lines in the system we were previously working.

Our field inspection yielded several interesting points. First, the portions of the line now replaced with FBE were demonstrated to have exceptionally high AC potentials. The maximum AC voltages observed on this line were double that shown on the downstream section and exceeding the “Safe Touch” potential of 15 Vac. This was not observed on the portion of the line section that had not been renewed; suggesting more of the poor coating conditions as previously observed on pipeline of this vintage.

Figure 1: AC Profile

To complicate matters, the line section was shown to fail protection criteria with subsequent increases in protective current. Potentials on the new line were shown to rise accordingly, though resulting in potentials known to damage FBE type coating systems. This demonstrated a distinct imbalance between the two portions of the same line in terms of resistance to ground. The existing impressed current system for this line was also at or near it’s maximum rated output, and so more current was as infeasible as it was impractical and damaging for the new coating system.

We returned to our client to recommend that the line section be divided into two parts, effectively isolating the existing pipeline from the newer line for a slew of benefits. The measure would prevent an active galvanic cell from forming should the existing impressed system be disabled. Additionally, it reduced the overall protection current required for the aged line as half of the system could now be removed. The new line could now be protected independently, and likely with a rather low demand in terms of protective current. Through our subsequent testing phase, we determined the specific properties of the line and prototyped protection and mitigation strategies.

AC Mitigation:

With regards to the AC properties, we found that the AC could safely be drained to the retired 12” main that now existed parallel to the live main. To arrive at this solution, we systematically tested drain points to establish the best location and ground structure along the line by charting AC current drain points to existing test points along the route. The highest discharged currents were then tested in test configurations by retesting each point with a ground strategy employed. We found that the AC was best mitigated through an AC drain placed at the midpoint of the inductance corridor attached to the retired main that had been bypassed.

A direct bond to this structure was not suitable as it would greatly increase the current requirements for the structure. Fortunately, the two properties can be treated differently through the use of a capacitor in the circuit. A capacitance of approximately 1mF or better was shown to have the desired effect, reducing the AC potentials to well below any hazard conditions. This was prototyped by wiring a 1 mF capacitor into a test station with a bond to the retired line. We then retested the line with this device in place to show that both AC potentials were sufficiently reduced while leaving polarized potentials in the desired range for protection.

Designs were then constructed to install a Solid State Decoupler (SSD) at the test point previously prototyped. Commercially available devices of this type were specified as they exhibit those properties necessary to achieve the necessary AC impedance and resulting current. They have the added benefit of providing an over-protection feature to prevent potentials to exceed those noted before as potentially damaging to the coating type. Not to mention, such devices were readily available to the client as they stock the devices for similar applications. The device could be mounted directly to the fink-type test station previously referenced.

CP System Design

The low current requirement of the system was in a range expected to be suitable for a passive, or galvanic, protection system. We set out to evaluate the area by collecting soil resistance measurements in the area using the Wenner 4-pin method and evaluating layer conditions using the Barnes’ Layer Analysis. Our analysis of the area determined that the soil resistance was exceptionally high, as was common in our experience in the area. The impact to the design was that it would require up to 4 separate galvanic groundbeds in order to achieve the desired output. Alternatively, we chose to submit a design for an impressed current system; whereby we were able to achieve the desired output with minimal installation and only 4 anodes placed vertically.

Figure 2: CP Design Strategy

 

Design Complications:

We then began evaluating the potential to include some under protected station piping, at the request of our client. The design for the ground bed coupled with the modest current requirement meant that a significant amount of capacity would be left on the standard model rectifier in use. Coincidentally, the regulator station selected as the location for the ground bed was recently repaired, resulting in some station piping being isolated from its previous CP source(s). It was therefore convenient to add these short lengths of pipeline to the protection system. We knew that we could balance the circuit and control current direction and magnitude to each respective line.

Figure 3: CP Balancing Model

However, similar to our concerns for electrically joining the new FBE line with older sections, we had to take precautions when adding these new lines to the CP system. Firstly, we had to evaluate the resistance to ground of each of the lines individually. From this we could build a theoretical model for the systems to operate, thereby balancing the resistance differential of the systems to balance the current distribution to each respective system. Additionally, we needed a way to control direction of current as the new steel would inherently be more active than the aged steel at the station. This would ensure galvanic cells do not form if the CP system were to go offline for any period of time, as well as prevent recirculating currents in an interrupt survey. Lastly, we needed to know the empirical current demand for each system individually. Current demand tests were then conducted on each line individually to establish current demand and resistance properties.

As we were conducting this battery of tests we noted some abnormal results. The resistance to ground for each of these lines was abnormally low. More so, the current demand for each line was shown to be exceptionally high. We began applying additional test methods to better understand the properties we were observing. Through application of coating holiday tests, such as ACVG and DCVG we noted some rather sizeable indications centered upon the valves inside and immediately outside of the yard.

Figure 4: AVCG Indications

The DCVG portion of the test became difficult to interpret. With the complex routing of the various lines the direction of gradients became difficult to map underground direction. We altered our test method slightly to allow a more agnostic view of the lines. Using an adjacent CP system rectifier, we applied a test current with a fast interrupt period. To reference current direction, we placed a remote reference well outside of the gradient which we determined by evaluating for remote earth condition. We then took measurements in a grid fashion about the station yard. The result was interpreted by magnitude of shift at each location within the yard. This was visualized by creating a contour of those shifts in a 3d format.

Figure 5: DCVG Surface Gradients

Figure 6: Results of CP Surface Gradient Indications

The information that this contour map provided allowed for a prioritization of the holidays previously indicated that could then be pursued. Based on this prioritization, several locations of the station were scheduled for excavation. The result was the discovery of bare sections of pipe and fittings within the station, resulting in lower resistance to ground and elevated current demands. These areas were repaired with wax tape coatings and backfilled.

The result of this test sequence was the addition of one of the three steel lines in the station to the circuit. One of the lines in the station was found to be shorted through its isolating joint to the adjacent CP system. The current requirement of this system made it infeasible to protect with the new system, and so it was recommended to evaluate it as part of the adjacent line section. The remaining 6” line transitioned to plastic at the adjacent road and increased the current load by only 60mA. This line was balanced against the high-pressure line by way of a 1ohm resistor, as determined in the circuit model of the two lines. The two lines could then be isolated on the DC, or time domain, with the insertion of a simple diode. The diode would act to isolate the two lines and prevent recirculating currents in the event of current interruption such as an unplanned power outage.

Commissioning of the CP system:

Following the installation of the CP components, we resumed work on this line to commission the new protection system and ensure its effective implementation. In our evaluation of the system we conducted both an interrupted survey of the line as well as depolarization survey in order to assess effective polarization levels on the line. In this sequence we realized that the line would not pass the desired criteria. To complicate matters, the line was found in some areas to exhibit “native” potentials more electronegative than the desired -900mV CSE criteria to be employed.

These points led us to evaluate the line for foreign or unintentionally applied protective currents. The locality of the highest “native” potentials coincided with some repairs in asphalt over the line. This led us to believe that a short to a service might be to blame as an intermediate pressure distribution line routes in parallel to the high-pressure line; servicing the houses in this neighborhood. We set out to test this hypothesis by way of some indirect inspection techniques: namely Current Mapping and Close Interval Survey. Both tests gave negative indications for anodes or foreign contacts.

Figure 7: CIS Results: In Search of Anodes/Shorts

As observed in our results, there was not definitive indication for an anode despite the variations in the profile. Of note: the peaks observed with the ON-OFF survey did not align to peaks observed in the Native profile. Should a directly connected exist there, we would expect to see the potentials spike in both profiles. Additionally, the ON-OFF would have demonstrated a lower resistance to ground with the increased current traveling to the effective coating defect. In the shown results, no major drops in the calculated resistance to ground (Rvg) were observed.

Next, we evaluated the line for indications of stray current. By systematically testing the line and sequencing adjacent or geographically close rectifiers, we were able to identify interference effects on the line from three separate rectifiers. Of the three rectifiers, the closest rectifier had the most impact as it influenced the native potentials on the line and demonstrated a discharge condition at the most cathodic points on the line.

We then set out to evaluate the effects of the three systems and determine if a new mitigation strategy was warranted. In order to observe a true depolarization, we shut off all 4 CP systems aforementioned for a period of 48 hours and resurveyed. Similarly, we energized all systems for a second 48-hour period. The line was resurveyed with synchronous interrupters installed at each of the implicated rectifiers. The outcome of this survey demonstrated the line to be effectively protected with sufficient polarized potentials observed in a fully interrupted condition.

Figure 8: Satisfactory Results

Conclusion:

This sequence represents the best cross section for complex designs. Many times, as field technicians we find that conditions change in the progress towards protecting our assets from corrosion. Recognizing the process is dynamic allows for adaptability in the analysis.

What started above as an AC evaluation led to a total redesign of the protective measures on a new system. Through subsequent tests, the line could be mitigated of its AC problems and tertiary systems could be added in close proximity. Finally, when all pieces arrive in place the outcome is not as expected. By applying the same open observations and corrosion principles, new conditions can be identified to allow for proper interpretation of these results.