Understanding the Risk-Based Inspection (RBI) Confidence Scoring System
Driven by API 580 and API 581 standards, the first Risk-Based Inspections began in 1994. The RBI program approached the quality of inspections as merely a function of the inspection time. The limiting factors of this approach were two-fold:
- There was no consideration for the value of inspection data density (i.e. ultrasonic measurements per square foot).
- And there was no consideration for the operational consequences of failure pertaining to a particular plant asset.
As the guidelines and technology advanced, the scoring methods for RBI became significantly more data centric and prioritized high risk assets that operate with elevated temperatures, pressures, damage mechanisms, and corrosive chemicals.
At its core, RBI scoring sought to create a standardized industry-wide approach to inspections. With an effective RBI scoring system lowering overall risk and optimizing inspection costs for even the most at-risk assets, technology focused on the efficient collection of massive amounts of data became more accessible.
The data-driven confidence scoring system outlined in API 581, and summarized in an article with Inspectioneering Journal aims to guide refinery RBI programs to achieve four goals:
- Identify and measure structural integrity risk for all plant assets
- Summarize and prioritize a thorough understanding of asset risks
- Enable effective risk management through preventative maintenance measures
- Reduce risks associated with facility operations
The scoring system summarized below outlines the RBI Scoring System:
The Problem With Trusting Low RBI Scores
For many refineries, budget restraints and outage induced time restrictions have forced facility management to reduce overall plant costs with cheap inspections. Many times, these low-cost inspections provide the minimum viable inspection coverage that corresponds to RBI Confidence scores of “C” or “D”. For a select few situations, this inspection level may be all that is required to proceed effectively. However, many more will find themselves in a situation similar to our example.
ON-SITE EXAMPLE: DONNIE
Picture Donnie as a plant manager for a refinery in Houston, TX. If Donnie contracted a particular inspection method that returns a clean report showing little corrosion but has a “C” score, the variability in inspection results of a “C” level score could mean that the asset being inspected has been partially inspected or totally un-inspected areas. An example of this would be traditional methods of gridded, manual UT inspections that typically have only one or two UT readings per grid square, typically 1"x1". These areas could contain corrosion that could compromise the structural integrity of the asset wall, yet not be identified by the aforementioned level of inspection. Without additional inspections returning a higher confidence score, Donnie can not continue to operate the asset in question with a significant level of confidence.
Because of this, Donnie will need to procure additional inspection(s) that return a higher confidence score and will be required to do so before continued operation or maintenance measures can begin. This could increase total inspection costs, not originally budgeted, as secondary inspections will be necessary to identify all areas of wall thinning.
Maximizing Inspection Confidence Scores with RUG
To address the deficit in confidence score, refinery leaders like Donnie are turning to Gecko Robotics for two reasons. Firstly, Gecko's Rapid Ultrasonic Gridding (RUG) inspections frequently are scored in RBI programs with A-ratings and have no scores lower than a B-rating. These scores give facilities the confidence to take decisive actions. In most situations, the decision is either to proceed with maintenance measures to resolve wall corrosion or to quickly return the inspected asset into operation. Secondly, Gecko's robotic inspections provide some of the fastest whole-asset coverage available on the market. Completing inspection coverage using RUG is vastly superior to other inspection methods that can return a comparable RBI Confidence score because RUG is 10 times faster than traditional gridding and competing methods like AUT.
These points are illustrated in an example inspection scenario comparing the inspection results of RUG and Automated Ultrasonic Testing (AUT) on four (4) bpx condensate separators. In the scenario, the bullet tanks being inspected were 12' diameter and 80' in length. With RUG, a 2-person crew could complete the inspection in 4-days, capturing roughly 7.3M A-scan measurements while reducing the risk of inspector injuries through ground operation of the inspection robot. The total reduction in cost for this job was 75% in comparison.
The other popular automated inspection method that consistently returns high level RBI Confidence Scores is Automated Ultrasonic Testing (AUT). Using AUT on the same set of tanks would require a 4-person crew, but would need 28-days to complete. Choosing RUG over AUT in this scenario would yield the same RBI confidence level in 1/7th the time and 1/4th the cost.
In summary, RUG can be a key inspection tool for maximizing value, speed, and inspection coverage to deliver high-level RBI Confidence scores at a cost-effective price point. In addition, leveraging Gecko’s RUG process includes inspection data storage and access via the Gecko Portal, which features color-coded Bokeh plots for swift identification of high corrosion areas of the asset under inspection.