One of the biggest challenges of the oil and gas industry is the development and implementation of new technologies to enhance recovery from mature fields and extend the life of ageing facilities. Asset owners seek more effective methods for managing their existing assets and future developments, increasing performance and reducing costs.

However, how does one quantify the performance optimization gained by the application of these methods and from the implementation of new technologies on a mature asset?

RAM Analysis

A proven methodology named Reliability, Availability and Maintainability (RAM) analysis is the answer. Through RAM analysis it is possible to analyse and improve the production efficiency of assets, safely and responsibly.

The first step to evaluating opportunities for production optimization on a mature asset is building a detailed model that covers the design of the facility, the operational procedures and maintenance strategies. The outputs for the study will:

• Provide a quantitative picture of the asset’s performance over its remaining life
• Identifycritical equipment and systems, providing focus to improve the asset’s performance
• Quantify the impact and/or threats from potential investments

The analysis will:

• Give you a better understanding of performance behaviour, describing pertinentparameters such as produced volume, availability, maintenance, resource use and operational constraints
• Identify which areas requirefocus on performance improvement initiatives
• Rank initiatives based upon the greatest return on investment
• Improve performance safely and responsibly
• Give a full picture of annual production efficiency over the asset’s life

Which Parameters are Considered?

A RAM study in the oil and gas industry needs to include all essential capabilities such as:

• Failure distribution for all equipment
• Repair distribution for all equipment
• Ability to model logical events to represent operational strategies or constraints
• Various planned intervention strategies
• Impact of seasonal constraints in regards to intervention strategies
• Export constraints, for example, seasonal issues, number of vessels, etc.
• Individual well production profiles, including a demand profile
• Wells phasing in and out over the asset’s life
• Resources availability such as vessels, spare equipment and crews
• Economic parameters (mobilization costs, daily cost for operations, oil price, etc.)

Advanced RAM Analysis

By combining experience ofoperational procedures with advanced RAM analysis, a simulation model that accurately reflects the failure patterns and proposed operating strategyfor the asset is achieved. This leads to accurate results for the following performance parameters on an annual basis throughout the field life:

• Achieved production efficiency and production losses
• Rank of critical systems, equipment and modes of failure
• Detailed result for the number of vessel interventions taking into account mobilisation time and hours of usage broken down by activity
• Operational costs, revenue losses and through-life NPV reduction due to subsea failures/activities

During the early stage of the project, the use of this approach ensures that competing technologies are evaluated, focusing on different results in predicted performance. Comparison of predicted through-life net present value (including incremental capital costs for any additional equipment for each option) allows selection of the optimum configuration or operating strategy.

Dynamic Simulation Techniques

The traditional approach of using static calculations to determine operating costs and production losses for assets in the oil and gas development produces a generalpictureof the cost through the asset’s life. However, use of dynamic simulation techniques, calculations measuring continuous changes in the state of the system over its expected life, ensures that the predicted performance includes the impact of the operating environment, its constraints and any potential changes over field life.

Incorporation of all these factors in the analysis results in more realistic performance figures and improves confidence in the predicted values and, hence, in any derived recommendations. One of the major benefits of advanced RAM over traditional methods is the ability to incorporate your understanding of the asset and the way it operates within your organization.

Each asset and operator typically has specific issues, so the use of an across-the-board standard solution is often not advisable.

Case

Normally Unmanned Installation (NUI)

A normally unattended installation,also referred to as a normally unmanned installation,is an offshore installationdesigned to be operated remotely through automated processes and without the presence of personnel. NUIs are commonly utilized in shallower water where building many small NUIs is an easier and cheaper option in comparison to the cost of using subsea wells. NUIs can also be used as a support platform to a nearby larger platform.

This type of platform requires a unique maintenance strategy and its effectiveness can be heavily influenced by several parameters such as travel times, maintenance resource constraints, mobilization time and prioritization of repair. Advanced RAM analysis allows you to integrate all of these factors into a performance prediction study. By running an advanced RAM analysis, the user will be able to check the effectiveness of a maintenance strategy and the criticality of the different systems comprising a platform.

In order to get a complete overview of the performance of this specific asset an advanced RAM tool must be used.

For this particular example, DNV Software’sMaros was used.

Maros, which stands for Maintainability, Availability, Reliability, Operability, Simulation, is an advanced RAM tool with extensive features for modelling networks, maintenance, operations and demand scenarios which includes powerful Boolean logic option.

The objective of the study is to analyse the production efficiency and availability for a specific NUI configuration and maintenance strategy. The goal is to understand the key contributors to the loss of production and to analyse the viability of several potential operational changes.

The Modeled System

The model consists of an NUI with four oil wells and onewater injection system.

The platform is at the end of its life –the assumption is that operations started 10 years ago.All flows coming from the wells converge to the NUI platform passing through a water injection system. All wells have a similar system including valves and tubing. A planned inspection of the subsurface safety valve is carried out every sixmonths. The NUI comprises of different systems such as separation system, seawater system, telecommunication system, and power generation unit. Every time there is a failure leading to complete shutdown, the NUI platform must be restarted manually.

The maintenance crew flies by helicopter to repair failures. There is unscheduled maintenance to address production critical failures and a scheduled weekly maintenance visits to address non-critical failures.

Cost data is provided for:

• Scheduled helicopter visit $12,000.
• Unscheduled helicopter visit $15,000.
• Rig day rate: $60,000, diver support vessel day rate $45,000.
• Oil price: $20/bbl.

Modelling Considerations

Flow Profile

The ability to model the well deployment over the life of the system has a significant effect to the performance of a system and it must be accounted for to get a complete picture of performance. With a flow calculation method, the analyst is able to go beyond normal measure of availability and, instead of giving only an Availability result (system uptime), you can produce an estimated production efficiency.

By modelling the flow through the systems, the study takes into account degraded failures (partial loss of production) and other Operability factors which is not possible in traditional RAM Analysis. Degraded failure may account for a huge amount of losses. Hence, rather than just providing rudimentary uptime v.s. downtime information, production efficiency keeps track of how much production losses are throughout the system life, and quantifies the efficiency by dividing the actual production by the potential production. Combined with time varying flow from multiple sources, this result becomes a very powerful metric.

The expected flow of oil from each well is 12,000 bbls per day.

Events

In oil and gas developments, events can be separated mainly into three categories: unscheduled, scheduled and conditional. These three basic events are defined below:

• Unscheduled events are unplanned and occur at random.However, their occurrence usually corresponds to a particular statistical distribution. Example: equipment failures
• Scheduled events where the occurrence is known. Example: routine inspection
• Conditional events that are initiated by the occurrence of other events via a Boolean logic expression. Example: warm-up of equipment in standby

For example, the NUI has two of the three events defined under its system. One unit might comprise of only equipment failures, planned maintenance or conditional element or even a combination of these three.

The model is set up based on the Reliability Block Diagram (RBD), Figure 1 which will at times appear to be different from the actual process flow diagram (PFD) simply because RBD is built with focus on how the reliability of one system affects another system.

RBDs are used to outline the logical relationships between the different events, equipment and systems. The level of detail in these diagrams allows identification of relative equipment capacities and areas of operational redundancy.

Figure 1: Reliability Block Diagram

Each unit will comprise its own reliability data – for example, the data extracted from the separation system is shown as follows:

Maintenance Resources Logistics

An extensive number of maintenance resources must be accounted for when performing a RAM study. It is important to understand how these logistics manifest themselves in the simulation. Consider a generic event, a failure or a planned shutdown. In real life, this failure will start being repaired with a certain delay time corresponding to the time required to diagnose the problem and organize the repair resources to carry out necessary repairs.

Once all resources are on the job location, the actual repair procedure can commence. The repair action itself may impact the production in another manner.

Modelling maintenance logistics involves determining the 'repair delay' portion of the above sequence. This is achieved by defining the location, quantity and constraints of the various resources involved in the repair process. Simulation is then carried out to determine each repair delay depending upon the foregoing and the workload at the instant of the failure. It is not simply a case of specifying a repair delay per task.

Financial data

If the analyst has access to the values of capital expenditure, operational expenditure (cost for spare parts, man-hours, etc.) and product pricing (selling oil and gas,for example) a RAM study can be used to calculate the net present value (NPV) for each design possibility.

Calculation must take into account multiple products pricing which could be variable during the lifecycle.

Analysing the results

The analysisproduces a large range of results

The production efficiency of the system is 94.03%. The declining production profile can be seen Figure 2.

Figure 2: Production profile

Based on the subsystem criticality (Figure 3), the most critical system is the seawater system, which is responsible for 38.309% of the losses.

Figure 3: Subsystem criticality

Drilling into more detail, one level below the seawater system, the lift pump is shown as the most critical system, representing 36.5% of the 38.3% of losses.

The lift pump,as part of the seawater system, can be drilled-down into even further details.

The economic analysis gives the NPV of the project, and is summarized. In this case, there is a loss of $91 million throughout the life of the system.

Performing a ‘What-if’ Scenario

The results above provide a baseline performance for the normally unattended installations under study. Advanced RAM analysis allows the analyst to simulate and inspect the behaviour of a complex system under some given hypotheses, called scenarios. These scenarios used in theadvanced RAM analysis might include changes to design, maintenance strategy or operational procedure. The analysis is carried out by varying different parameters in a model such as reliability of certain equipment and/or number of crews or spares.

For example,what if there is a redundant lift pump in parallel dealing with 100% of the flow?The production efficiency of the system is now 95.69%, an increase of 1.65% over the base case.

Conclusion

The typical key results from a RAM study include a better understanding of system performance and criticality.

System Performance

Performing a sensitivity analysis allowed us to predict the production efficiency for different design options:

Not only providing the user with the performance (production efficiency) of the asset, the analysis can provide one step further analysis into the cost benefit analysis of the possible investment project. In this case, by looking at the sensitivity case 1, the investment is likely to be viable.

Criticality

Criticality and key problem areas

• A list of the most critical equipment and systems is produced by the end of the analysis
• The seawater systemrepresents38.31% of all the losses registered for the base case
• The maintenance campaignrepresents20.42% of all the losses registered for the base case
• Host platform trips represent 14.96% of all the losses registered for the base case

A large set of variables could be changed aiming at a better design and performance, effective maintenance operations and reduction of operational constraints.

A RAM study will support decision-making that will optimize productions at facilities in a safe and responsible way. A software package makes it possible to efficiently take into account all these key parameters.