Reliability analysis points to $500,000 savings

The use of reliability analysis by engineers at FMC Energy Systems pointed the way towards $500,000 savings in a hydraulic power unit (HPU) used to power a subsea production system. Reliability is among the most important design issues in hydrospace due to the difficulty of repairs and rig downtime costs in the order of $25,000 per hour.

On a recent production system, FMC engineers originally planned to implement dual, redundant HPUs so that the second unit could take over in the case of a failure of the first.

In order to validate their approach, reliability block diagrams were used to model the interdependencies of logical system components and compute the overall system reliability, while taking account of various redundancies such as active, standby, m-out-of-n, etc.

The analysis showed that the reliability of the overall production system with a redundant HPU was slightly higher in the early phases, but converged quickly with a less expensive, single-HPU approach, which was also easier to maintain and took up less space. As a result, engineers proposed a design using a single HPU with spares available for quick changeover, providing substantial cost savings with minimal reduction in reliability.

FMC’s strategy of developing standardized subsea production components and systems offers customers advanced technology with lower capital expenditures, faster system deliveries and reduced operating costs.

“As the subsea industry moves into deeper waters, technical challenges and associated risks increase and failure costs rise at a rapid rate,” said Frank Wabnitz, reliability engineer for FMC. “Yet the complexity of the equipment used has increased to the point that determining the reliability of various design alternatives is a very difficult computational challenge.

“To address these issues, we developed reliability engineering expertise internally to provide assistance to engineering,” said Wabnitz. “One of the projects that we have supported was an HPU that sits on a rig and provides hydraulic power through umbilicals to the subsea Xmas trees and manifolds that control the flow of oil and gas from the well.

“The basic issue was balancing the tradeoff between reliability and cost, size and maintenance requirements of the unit. We wanted to maximize reliability, yet we recognized that adding redundancies to that area might drive up cost and complexity while having little impact on overall system reliability. What we needed was a way to model the overall reliability of the system while comparing various design alternatives.”

On previous projects, Wabnitz identified reliability block diagrams (RBD) methodology as a valuable analysis tool. A RBD represents a clear picture of system functions and interfaces and shows the failure logic of a given system. It models the interdependencies of logical system components and allows computation of the overall system reliability.

Once a model is established, sensitivity analyses can be performed by changing configurations, adding redundancies and modifying maintenance strategies, to name a few options.

Developing a RBD using pencil and paper or a spreadsheet is a long and involved process that is fraught with the potential for error. To streamline the process, Wabnitz looked for a software package that would simplify and streamline the process while providing computational tools to improve its accuracy. He selected Relex RBD from Relex Software Corp., Greensburg, Penn.

With this software package, an engineer simply needs to place the figures on the screen, assign properties and connect them together to generate a RBD analysis. The software then provides fast and accurate reliability, availability, mean time between failure (MTBF) rates and failure rate calculations through analytical methods or a built-in Monte Carlo engine.

“A key reason for selecting Relex was that this package incorporates all the important reliability tools, reliability prediction analysis, failure mode and effects analysis, fault tree analysis, maintainability prediction, life cycle cost and system reliability analysis in a single integrated system,” Wabnitz said.

Two senior design engineers at FMC, responsible for designing the HPU, were supported by Wabnitz in their efforts to provide an optimal, reliable solution. “We started by reviewing the system requirements and architecture and identified potential failure modes for each item,” said one of the engineers.

“Next we considered all the interdependencies in the system. An example is a series configuration — if one component fails then the entire system fails. We selected the appropriate figures from the toolbar and dragged them into the page to create the diagram.

“The next step was determining failure data for the individual components. It’s difficult to obtain this type of data, even from our customers, because the subsea production business tends to be quite secretive. We obtained reliability information for most of the components in the HPU from the NPRD95 database.”

Once the model was completed, FMC’s engineers used the Monte Carlo simulator to analyze the diagram. The Monte Carlo method simulated the life of thousands of virtual components while considering the failure probability of each to make a determination of how and when the overall system failed. The entire population of components was then statistically analyzed to determine all of the key reliability statistics as well as which failure methods predominate.

The model was changed and recomputed for a number of different design options revolving around the possible use of a redundant HPU. The results showed that the option with dual, redundant HPUs provided higher reliability in the early part of the HPU’s lifecycle but that both options converged relatively quickly and from that point on delivered similar failure rates and reliabilities.

The option of using the single HPU met the customer’s MTBF requirements at a substantially lower initial cost while also providing space and maintenance savings. As a result, the engineers proposed a single HPU to the customer, with spares provided for quick changeover in the event of a failure. “It’s clear that this approach will provide substantial cost savings without a major impact on reliability,” one of the engineers concluded.

Based on this and a number of other successes, FMC managers made the decision to implement reliability analysis to identify and reduce risk during the early development stages.

“Reliability engineering resources were added and a global reliability program was developed with the dual objectives of optimizing performance and minimizing overall system life cycle cost,” Wabnitz said.

“We have overcome a major roadblock, the lack of numerical reliability data, by using risk assessment tools that deliver a wealth of numerical data.” The benefits seen include:

1. Clearly reduced development cycles that identify risks early and the need for costly redesign cycles during the later phases

2. A reduction in development and production cost, and

3. Increased reliability resulting in lower operating costs.

FMC Energy Systems is based in Houston, Tex. For more information, contact Relex Software Corp., 540 Pellis Rd., Greensburg, PA 15601-4584, tel. 724-836-8800, e-mail info@relexsoftware.com, or visit www.relexsoftware.com.