ICDA Successfully Applied for a Piggable Refined Product Pipeline
Ashish Khera, P. Eng
NACE Instructor
- Direct Assessment
Allied Engineers,
New Delhi, India
Bidyut B. Baniah
Allied Engineers,
New Delhi, India

As per laws of nature, iron/iron alloy such as carbon steel used in manufacturing pipelines will eventually want to return to its natural form. Based on the in-place, asset integrity management practices of the pipeline owner this process can be delayed and ideally controlled.

Learning from the recent catastrophic gas pipeline failure incident in India , an Indian pipeline operator with a Trans-India presence of refined product pipelines decided to take the high road and have their jetty pipelines assessed in the year 2015. Majority of the pipelines under the scope of assessment were termed as 'non-piggable', whilst the remaining 20 per cent that is three pipelines was classified as 'piggable'.

Using OISD's guideline document (no.233) for inspection of non-piggable pipelines as the base, the operator took complete support from NACE International's Direct Assessment (DA) standards to design their scope of work. Using the NACE DA Standards as well as OISD's Standard Operating Procedure(SOP) released in November 2014 - the operator targeted to complete the assessment of these pipelines to define the further course of action.

What is Direct Assessment (DA)?

The four (4) step, continuously iterative process requires the integration of data from available histories, multiple above ground field surveys, internal corrosion predictive modeling (ICPM), detailed examination on the pipe surface and finally concluding the whole process in the fourth and final step known as the Post Assessment stage. Typically, DA process combines both existing and newer techniques for understanding the integrity of the pipeline.























Figure 1: Extensive debris retrieved on cleaning pigging

The key major advantages of DA over the other available integrity assessment techniques are:
  • Can locate and predict susceptible areas where corrosion could form in the future rather than areas where defects have already formed (proactive)
  • No product interruption is required(non-intrusive)
  • Mandatory for DA programs to perform 'root cause analysis' for determining the corrosion mechanism. Answers the principal question asked by operators 'Why does the corrosion anomaly exist?'
  • A need to deliver a go-forward plan(mitigation plan)
These advantages highlight the major technical differences between Intelligent Pigging, Pressure Testing and DA and this is how these techniques can complement each other rather than compete each other. The fact that 80% of the pipelines under the scope of assessment were not designed to be piggable, altogether removes the option of Intelligent Pigging with conventional Intelligent Pigging tools. Although, today there exists highly advanced high fidelity intelligent pigging tools designed specifically for non-piggable pipelines. At the time of the job execution, the operator preferred to utilize the option of performing Direct Assessment (DA), specifically External Corrosion Direct Assessment (ECDA) and Internal Corrosion Direct Assessment (ICDA).


















Figure 2: Predicted corrosion rates for the pipeline without oxygen ingress

Focus

Although, the integrity assessment for all fifteen (15) lines under the scope of study was deemed successful. The objective of this article is to focus on the results obtained for the time dependent threat of internal corrosion (using ICDA) for the three (3) refined product pipelines under the scope of study which were designed to be piggable. These lines also undergo routine maintenance pigging being performed by the operator.

It further discusses the advantages and the complexities to allow for a successful ICDA program. Even though, the three (3) pipelines under discussion may be relatively new as commissioned in year 2007, but lack of an integrity management program or incorrect operation practices leads to missing data, procedures or gaps in historic information. This critical information greatly assists in performing an effective internal corrosion integrity assessment.

It is hoped that petroleum pipeline operators use this as an example to further educate the industry to proactively assess corrosion, risk-rank it and then effectively mitigate the cause in a sensible economic and environmentally friendly manner with public safety being of paramount importance.

Steps-1 (Pre-assessment) and 2 (Indirect Inspection) of DA

Extensive historical data collection was executed during this stage. Pipeline operator personnel from the technicians in the field executing the tasks to the managers were interviewed in detail. Based on the preliminary data collected and discussions, it was felt by all parties involved that internal corrosion threat should be minimal for such lines because:
  • Relatively new lines (2007 commissioning)
  • Refined product being transported(less propensity for carrying any dissolved gases and water which are the key drivers for internal corrosion)
  • Routine cleaning pigging program in place
  • Confidence of the operator for the overall health of these pipelines
Subsequently, a cleaning pigging was planned in the presence of ICDA team. The main idea was to collect solid and liquid samples (if any) along with advanced lab testing. For pipeline of only 7 Km length, the expectation of any experienced pipeliner would be that of a few Kg’s of debris/muck. The results of cleaning pigging were opposite! (refer Figure-1)

For two (2) out of the three (3) pipelines, the volume of debris retrieved could be quantified at approx. 100 Kgs. This was not the case for the third (3rd) line, i.e. the HSD product line. This line also so happens to be the pipeline that is under operation with maximum frequency that is 22 days per month. Coincidentally, the other two (2) pipelines of MS and SKO product were operated for a time period that may only equal to 8 days and 3 days per month respectively.

To deduce the above situation, the MS and SKO pipelines remain under 'shutin'/dead-leg condition for 73 and 90 percent respectively!

The liquid samples collected were further sent for wet chemistry lab testing . Balance water samples extracted from tank bottoms as well as the pig receiver were further analyzed for DNA genetics bacteriological tests.

The solid samples collected were further sent for the following tests:
  • Energy Dispersive X-Ray Analysis
  • X-ray Diffraction (XRD) Analysis
  • Scanning Electron Microscope(SEM) Analysis
Based on the unexpected results, a discussion was further held with the operators, team. This was when it was further divulged that the subject pipelines are flowing liquid (product) only when a ship is berthed. The cleaning pigs for the SKO and HSD product pipelines are pushed with compressed air, whilst for the MS pipeline, the pig is pushed by 50 percent nitrogen followed by a column of compressed air. Compressed nitrogen is used for the first half column of length in order to avoid any spark caused by ingress of oxygen from compressed air.




























Figure 3: Predicted wall loss % post re-modeling with oxygen ingress and other factors

Utilizing all the information, it was decided that the authors will perform multiple scenario simulations using proprietary internal corrosion predictive model called the 'Teevens Model' in order to capture the various operational scenarios that the subject pipelines have undergone throughout its operational history. 'Teevens Model' is commercially known as enpICDATM. The predicted uniform corrosion rates and subsequent remaining walls for the pipelines taking into consideration no ingress of oxygen in the pipeline system can be seen in Figure-2.

For the same system, if we consider the below factors for conducting ICPM:
  • - O2 @ 0.5 mol% (due to incorrect operational practice of utilizing compressed air which may remain shut-in within the system until the next ship/vessel berths)
  • - CO2 @ 350 ppm
  • - H2S @ 0.1 ppm
  • - Water flow rate @ 0.5% of the product volumetric flow rate
The subsequent wall loss % predicted post re-modeling in this case is shown in Figure-3.

Step-3: Actual in-field inspections:

From Figure-2 above, we can see that the internal corrosion predicted is very low in the range of 0.4 to 0.65 mpy (mils per year).

Post re-modeling using enpICDATM and modified conditions mimicking the field scenarios which can be otherwise termed as 'upsets'/improper operational practices etc. the predicted wall loss is illustrated in above Figure-3. This would mean a very high corrosion rate leading to loss of 50 per cent of the pipe wall in only 7 operational years!

Locations were selected based on the results of Internal Corrosion Predictive Modeling (ICPM) of water accumulation, solid accumulation and cumulative wall loss predicted for various scenarios. The pipeline was day lighted at these locations and as per below Figure-4 depicts an Ultrasonic (UT) based wall thickness measurement from the field.

The nominal wall thickness of this pipeline was 9.5 mm (0.375"). For an anomaly of remaining wall 5.48 mm. The calculated actual wall loss was 42.3 percent. The predicted wall loss based on ICPM was 50% which was performed theoretically by simulating the pipeline historic operations and getting the science right!

The difference between theoretical modeling calculations and actual field findings was a mere 7.7 percent!

Conclusions of the Program
  • a. Confirm presence and severity of the threat of internal corrosion non-intrusively utilizing proven 'Teevens Model'. This underscores the ability of a strong science and engineering basis to define pipeline integrity from first principles
  • b. Advanced liquid deposition modeling capability of the Teevens Model is proven
  • c. Advanced solid deposition modeling capability of the Teevens Model is proven. Solids deposition prediction leading to under deposit corrosion (UDC) and also erosion corrosion/abrasion
  • d. Allowed the operator to take immediate action of provision of corrosion inhibitor
  • e. Allowed the operator to take immediate action of executing an Intelligent Pigging program
  • f. Based on the nature of the internal corrosion observed by manual UT inspection i.e. narrow channeling corrosion as well as uniform corrosion, the Operator was advised to conduct Ultrasonic (UT) based "direct measurement" intelligent pigging
  • g. Operator decided to perform Magnetic Flux Leakage (MFL) intelligent pigging with both specifications of axial and circumferential direction based tools, which does remain an indirect inspection
  • h. Because of the inherent nature of the MFL technology as can be seen in Figure-5, it was not possible to identify the very narrow channeling corrosion anomalies as well as uniform corrosion due to low/nil flux leakage. As shown in below figure, an elongated anomaly is seen as 'spots' by a MFL tool, but with an UT tool the actual 'interaction' can be fully visualized and measured.
  • i. Unfortunately the Intelligent pigging report provided by the ILI vendor was not able to catch the elongated internal corrosion wall losses that was reported and verified during the ICDA program




















Figure 4: measured wall thickness from field





















Figure 5: (Top) corrosion anomaly; (middle) how a MFL tool sees this anomaly; (bottom) how an UT tool sees the same anomaly

The Operator did utilize the predictive based Internal Corrosion Direct Assessment (ICDA) as per OISD standard 233 and respective NACE International standards, to know that the relatively newer product line was suffering through severe internal corrosion. Intelligent pigging was followed and gaps due to the MFL technology limitations compared to Ultrasonic based In Line Inspection were evident for identifying the severity of the anomalies. ICDA and ILI do remain as tools available for the pipeline owner to assess the integrity of their asset and be able to manage the corrosion!