RBI Deviations From API RP 580/581

ReliaSoft’s RBI software facilitates risk based inspection (RBI) analysis for oil & gas, chemical and power plants in adherence to the principles and guidelines presented in the American Petroleum Institute's recommendations in the API RP 580 and RP 581 publications, as well as the American Society of Mechanical Engineers' recommendations in the ASME PCC-3-2007 publication.

The below sections contain deviations between ReliaSoft’s RBI software standards and API RP 580/581 Second Edition September 2008. Deviations are grouped by the part of API RP 580/581 in that they relate.

Part 1 Assumptions/Deviations

1. ReliaSoft added area consequence to PRDs when failure to open occurs, since resulting issues with over pressurized components can cause area consequences.

2. Planned replacement cost (CostPlan) should not have environmental cost, as there should not be environmental damage when a planned replacement occurs.

Part 2 Assumptions/Deviations

1. Probability of failure (Pf(t)) is calculated using P(t) = 1- exp (-gff*D(t)*FMS) as opposed to just P(t) = gff*D(t)*FMS (eq. 2.1). The reason behind this comes from the understanding that the non-exponential equation is just an approximation for the exponential form of the equation, however, when gff*D(t)*FMS gets larger, the error increases to the point where the Pf(t) can end up being greater than 1, which should not be possible.

2. Eq. 2.14 does not adjust for the possibility that the cladding has not all corroded away. This way the (age – age_rc) can become negative which leads to the addition of thickness via base metal corrosion in the equation. Fixed with max (age-age_rc,0) for C_r,bm and max(age_rc,age) for C_r,cm.

3. Based on the damage factor inputs, the plan date is also tested against the corrosion rates, so that the user cannot expect a plan date that would put the component where the wall thickness is less than the minimum wall thickness. The Plan Date will be automatically adjusted to reflect this date where according to the estimates minimum wall thickness is reached.

Part 3 Assumptions/Deviations

1. Fixed Table 5.2m, several choices had incorrect boiling points, ambient states, Cp values/equations. Used NIST database to validate/change values as needed.

2. Eq. 3.31, 3.37, 3.43, 3.49, are missing a 929 multiplier from next to the C7 as the area limit is stated in 3.5.8.3 to be 929 m[math]\displaystyle{ ^2 }[/math] and not 1 m[math]\displaystyle{ ^2 }[/math].

3. Area consequence due to flammability are often smaller for personnel than equipment, which is counter intuitive as the threshold limits specified in 5.8.2.3 show personnel limits to be lower than equipment.

4. The value for d in instantaneous release for HF in Table 3.5.11 is written as 33606, which probably should be 3.3606 and is treated as such. With 33606, the consequence area becomes 10^~33606 m[math]\displaystyle{ ^2 }[/math]…..

5. Several other Tables/Equations had smaller typos, or incorrect/incomplete data.