TECHNOLOGY BLOWOUT RISKS CUT WITH CONTINGENCY PLAN

L. William Abel
Wild Well Control Inc.
Spring, Tex.

A properly constructed blowout contingency plan can improve operator response time in a well control emergency, minimizing loss and additional risks.

A general blowout contingency plan should be developed before a well control event occurs, not during the problem.

Many operators think that each blowout is unique and that all techniques required are one-of-a-kind or applicable only to that event. However, many of the techniques, of equipment, and services are common to blowouts in general.

No matter how well a drilling or production plan is designed to prevent the loss of well control, the chance of blowout will exist. The question during planning should be, "When and how severe will be the blowout?"

Fortunately, major blowouts are fairly infrequent. But this infrequency leaves the operator's staff inexperienced in controlling blowouts; therefore, the operators generally rely on outside experts who regularly handle these problems.

The increase in number of high pressure, high temperature, deep water, and high technology wells challenges the technology level of well control in most companies. The pre-engineering and planning for regaining control falls clearly on the operating company, which is ultimately responsible.

A well control effort places a tremendous demand on a company's manpower, equipment, and finances. Most corporations maintain risk management positions so that large portions of the costs will be covered by insurance. In some cases, however, considerable time may pass before the actual losses are reimbursed. Thus, the company will have to provide the funds and logistical support to control the well; in major events this amount can be quite substantial.

The costs of control or the losses associated with fires and explosions in production operations are difficult to document because this information is typically not freely distributed in the public domain. Table 1 lists several large well control events and the approximate loss amounts. The amounts are gross approximations excluding reserves losses.

Given the cost of controlling a well, it is reasonable to commit engineering and planning efforts to create a plan for control. The result should reduce the cost of killing a blowout well.

CONTINGENCY PLAN

• The definition of scope, including the criteria for determining the severity of the problem and a scenario classification to quantify the severity

• The development of the contingency plan, including the possible solutions to the problem (intervention and kill techniques), the organization structure (primary response team and project management overview), and resources.

SCOPE

Defining the scope of the operational plan is one of the most difficult tasks because it requires forecasting. Because of the possible ramifications, the scope of the project must be fairly accurate. The range of situations can vary from a minor event in an easily accessible unpopulated area to a catastrophic worst-case scenario in a populated or poorly accessible area.

The blowout contingency plan generally cannot effectively encompass all areas, situations, and conditions. If significant differences in these factors exist, it may be prudent to develop individual plans tailored for each unique requirement.

The following outlines the various factors to be considered in defining the project's scope.

The initial consideration is the geographic region covered by the plan. Several operators have tried to create a worldwide, all-encompassing plan. This strategy may not be advisable because conditions vary widely and the plan will be too burdensome and complex to be useful. The geographical region should be carefully researched so a feasible workable plan is produced.

The potential downhole hazards of the specified region must be determined. These can include the following:

• Shallow gas

• Abnormal pressure and temperature gradients

• Abnormal fracture gradients

• Extensive reservoirs with high permeability and deliverability

• Hazardous fluids (H2S, CO2, etc.)

• Drilling hazards conducive to loss of control (lost circulation, etc.).

These hazards are not necessarily high risk factors associated with the possibility of a blowout, but merely factors that can make blowout control more difficult. Further consideration must be given to the proximity of the location to populated and environmentally sensitive areas.

LEVELS OF SEVERITY

The various types of losses during a well control situation directly influence how decisions are made. The blowout contingency plan must address the major types of damages or losses: human life or injury, equipment, hydrocarbon reserves, pollution, operational funds, public image, and the rights to drill and produce.

The classification of blowouts by type and degree of severity will yield plans matched to the event for more efficient operations. The goal of the blowout contingency plan must be to apply all necessary resources but without overkill or waste. The following is a general classification of well control scenarios:

• Class I-A minor event in which the well may only be leaking and is not on fire. Minor pollution might occur, and hazards are minimal providing the condition remains stable and other failures do not worsen the situation.

• Class II-A small-to-medium event in which flow rates range about 20-50 MMscfd of gas and 20-30,000 bid of liquid. The flow may exit either subsurface or above the surface of the seabed. The well is not on fire, and access to the wellhead is possible. Pollution may occur but is not major, and the fluid is not considered toxic.

• Class III-A small-to-medium event in which flow rates range about 20-50 MMscfd of gas and 20-50,000 bid of liquid. The flow may exit either subsurface or above the surface of the seabed. The well may or may not be on fire, and access to the wellhead is not difficult. Pollution may occur, and the fluid can be hazardous.

• Class IV-A medium event in which flow rates range about 50-100 MMscfd of gas and 20-50,000 bid of liquid. The flow may exit either subsurface or above the surface of the seabed. The well may or may not be on fire, and access to the wellhead is difficult but possible. Large amounts of pollution may occur, and the fluid can be hazardous.

• Class V-A major event with large flow rates in excess of 100 MMscfd of gas and 50,000 bid of liquid. The flow may exit either subsurface or above the surface of the seabed. The well may or may not be on fire. Usually, access to the wellhead is difficult or impossible, as in deep water or on a severely damaged or destroyed platform. This type of event can be extremely costly and time consuming. Pollution occurs, and the fluid can be hazardous.

INTERVENTION PLANS

After the scope of the blowout contingency plan is defined, a plan of action may be developed to counter the problems presented by each scenario classification. Given the well conditions under each scenario, a specific approach may be established and the various methods of solving possible problems may be developed.

The following are the four basic approaches to a well control problem:

• Surface intervention and pump to kill

• Relief well and pump to kill

• Combination of relief well and wellhead intervention

• Unique (infrequent) solutions.

Once the severity of the well control problem has been determined (by reservoir models or by judgment and experience), a kill plan can be formulated. The material, services, and logistics required can then be determined.

The operator must then determine what type of kill is possible. For example, a problem on a well in deep water or where the platform was destroyed generally precludes the possibility of surface intervention, leaving the relief well scenario the only suitable method.

SURFACE INTERVENTION

Many specific details are needed if surface intervention is attempted. These problems generally are subcontracted to special companies that offer a range of services in surface intervention techniques. However, the operator should become acquainted with these techniques and the logistics required.

The following is a partial list of surface intervention techniques:

• Fire suppression and extinguishing methods (water, chemical, explosives)

• Severing techniques (abrasive, water jets, sawing, die cutters, and explosives) for casings, wellheads, and structural members

• Wellhead and tree removal and replacement under pressure (commonly referred to as capping operations, tree snubbing)

• Diversion of large flows containing abrasive and corrosive fluids

• Freezing techniques and hot tapping

• Snubbing operations.

RELIEF WELL

The use of a relief well as a well control technique basically involves establishing direct communication with the problem well by drilling a hole directionally to a specific downhole location in very close proximity to the problem well. Once communication is established, it should be possible to pump the well dead. Fig. 1 outlines the major decision process required for relief well planning.

Under certain well conditions, it is conceivable that both intervention methods would be required to kill a well. The combination of methods would most likely be necessary for the most severe situations.

UNIQUE SOLUTIONS

Occasionally, unconventional techniques may be more suitable. The extent of these methods may challenge the imagination or be as simple as allowing the well to die or bridge off on its own.

After the intervention process, a kill operation is normally undertaken. The type of kill must be carefully chosen to prevent further damage and risks. There have been cases where intervention and or kill operations worsened the problem.

For example, an offshore project in Indonesia in 1985 began as a single blowout of a drilling well on a 12-well platform. A well control company recommended intervention by shutting in the well with the existing drilling blowout preventers (BOPs). Upon shut-in, the well breached the casing and blew out a short distance below the mud line. In less than 14 hr, a 90 m diameter by 150 m deep crater opened under the platform. The platform then collapsed in the crater, and several other wells blew out.

The surface intervention technique probably led to the platform collapse and exposed the operator to additional losses. Had a snubbing or other pumping job been used instead, the well might have been controlled without loss of the platform.

The operation to drill relief wells cost more than $50 million and lasted 14 months. The snubbing technique would have taken less time and would have been less costly.

In another example offshore Mexico, a bullhead technique was attempted to kill a well. This method caused the casing to burst just below the subsea BOPs. The well was contained prior to the casing burst, but afterwards a catastrophic oil spill resulted.

In another case, a shallow gas blowout offshore West Africa, a surface intervention from a relief well went awry. The relief well killed the blowout well using a dynamic technique with seawater as a kill fluid. The pumps were shut down because of darkness, and the well blew out again.

The next step was a massive pump job using heavy brine, but this proved disastrous. The open hole section in the relief well fractured back to the surface, and both the platform and the jack up rig were swallowed in a crater. This blowout occurred in 1978, and as of mid-1993, the well is still blowing a substantial amount of gas.

These examples illustrate that the operator must carefully choose and implement the kill operation. The type of kill operation must be fully investigated to assure the operator it will work without exposing the company to additional risks.

This article does not address all of the kill techniques possible. However, the primary techniques available are listed below with a brief description.

• Shut-in at the surface after capping the flow

  This technique uses wellhead equipment to stop the flow and contain any pressure exerted at the surface. The downhole conditions and equipment must be sufficient to contain the pressure and the water hammer effects from a sudden closing of the well. The final kill is then handled in much the same way a production well is killed.

• Volumetric control

  Volumetric control can be accomplished if circulation is not possible and the well can withstand the pressure and stress of being shut in at the surface. This technique involves pumping fluid in from the surface and bleeding back excess pressure after waiting for the well bore fluids and the kill fluid to exchange places. This method can also be used to handle gas migration while constant bottom hole pressure is maintained.

• Snub into the well with kill strings or equipment

  The snubbing technique can be used successfully to enter either a pressured or flowing well. In most cases where snubbing has been used in post-blowout work, a kill string is run in the well to a sufficient depth to kill the well with pumped fluids. In rare cases, packers have been used to stop flows below casing leaks.

• Diverting the flow

  This technique handles the flow at the wellhead and diverts it away through a vent line. Diverting the flow gives access to the wellhead so other techniques, such as snubbing, can be used. The diverted flow can be routed to a pipeline or production facility.

• Dynamic kill

  The dynamic kill uses hydrostatic and frictional pressure to overcome the reservoir pressure in the producing zone. This method usually requires massive pump rates and careful coordination. The dynamic kill is usually performed from a relief well and sometimes from a string in the blowing well.

• Minimum kill

  The minimum kill technique is a pumping method used at the moment a flowing or blowing well is closed in. The pump rates, types of fluids, and density of fluids pumped during the kill operation are carefully planned so minimum pressure is exerted downhole during the procedure.

  For example, the friction and hydrostatic losses can be planned so the well is killed while minimum pressure is maintained against the well bore. This technique is useful in situations where the integrity of the tubulars or wellhead equipment is unknown or in question.

• Momentum kill

  The momentum kill technique uses a fluid plug pumped down to overcome the upward motion of the blowout. This top-kill operation can be done from the surface and does not require re-entry into the well with a work string. However, this method may require exotic fluids and very high pump rates.

A knowledge of the flow rate and consistency of the fluid must be known to plan the job. Because this method has the tendency to create very high hydrostatic pressures and friction losses, it must be studied carefully to ensure that the downhole equipment can withstand the stresses.

An operator must have a basic understanding of these techniques before the blowout contingency plan can be created. Then, a basic plan for the control of any class blowout can be organized. Any necessary special equipment should be either purchased and placed on standby or contracted such that it can be obtained in a reasonable length of time.

PRIMARY RESPONSE TEAM

The severity of the problem will dictate the size of the staff and effort for the task. The control of a major staff effort will be quite difficult.

Project teams are best suited to handle difficult tasks. Some operators have designated an internal primary response team consisting of staff members from several disciplines within the organization. If an event occurs, the primary response team assesses the severity of the situation and takes appropriate action to control the well.

A common mistake made by operators in creating blowout response teams is to burden the operations staff with both the wild well responsibility and normal duties. This double duty can put undue stress on the group. Of course, the most efficient means of controlling any project is to encourage dedication to the task. People cannot devote full attention to two separate tasks.

A good operational practice is the use of a group whose sole task is to manage the control project. Small teams are usually more effective than large teams. A few good, qualified people are better than a large staff of inexperienced people.

One study of successful businesses revealed that "teams that consist of volunteers for a task of limited duration, and are set in their goals, are usually found to be more productive. The task force group found to be highly effective also has the following characteristics:"

• Limited number of members (10 or less)

• Reporting level and seniority proportional to the importance of the work

• Limited (short) duration of team existence

• Voluntary membership

• Team created rapidly and only when needed

• Swift and decisive follow up

• Informal documentation.

The most important factor is for the team to have an action bias, or willingness to try; sometimes a chaotic action is preferable to orderly inaction.

The primary response team must make the following major decisions soon after a well control event occurs:

• Is evacuation of the site or facility necessary to reduce the risk to personnel? This decision must follow a predetermined plan, and the authority to act should always be on site.

• The situation must be evaluated. Is a simple and quick solution available? What are the risks? How quickly will the situation deteriorate? Will the event become more severe or deteriorate over time?

• What is the level of the control effort (Class I, II, III, IV, or V)?

• What team members will be assigned?

• What experts are required?

GOALS

Once the initial assessment is completed, an operational plan must be implemented. The operational plan should have short range and long range goals. The short range goals are maintaining safety of personnel and the general population, evaluating the situation in real time, determining the deterioration rate for the situation, setting an immediate plan of action for control, controlling pollution to minimize the environmental impact, and handling any hazardous materials.

The long range goals focus on an ultimate solution, if a short term solution is not feasible or probable. The object of these goals is to help direct the control effort in the most effective and cost efficient manner. Prudent action requires a careful evaluation of all feasible methods with only the most suitable ones retained for further consideration.

Depending upon the severity of well conditions, such as the threat to human life and possibility of pollution, the list of kill techniques should be reduced to a single method except for the most severe cases. In severe cases, a maximum of three methods should be considered.

A primary control method should then be designated, and the response team should concentrate efforts on that method. For a severe problem, secondary methods may be developed and possibly pursued with the understanding they do not deter progress of the primary method. Should the primary method fail or well conditions change such that the secondary option is chosen, the response team should shift efforts to allow the earliest possible control of the well.

The long range goals should include safety of personnel and population, control of the well, recovery or permanent abandonment of the well, proper handling of hazardous materials, and keeping to a minimum pollution, costs, reserve losses, and damage to public image.

PROJECT MANAGEMENT

The well control effort is best handled by project management. This approach is designed to plan, organize, staff, direct, and control the effort. The roles vary for each discipline (reservoir engineering, drilling engineering, logistics, and the project management staff). The roles of each asset group must be determined and authority and responsibility must be established, especially for the project manager.

The organizational structure must include lines of authority, job descriptions, and responsibilities, with corresponding lines of communication. Fig. 2 illustrates the general responsibilities of the team. Such a setup is vital to the effectiveness of the team because it must be well informed of the current situation, the goals of and major decisions by the management team, the legal and governmental implications, and public relations.

For the team to be effective, each member must know well his responsibility and authority level. Otherwise, the members will be ineffective.

To control the project, a comprehensive decision tree is helpful and, in many cases, mandatory. Risk models are also useful in the decision process. A decision tree can be too comprehensive or can include events that have a low probability of occurring. Above all, the decision tree must be a practical model for likely and probable events. In no way can one accurately predict all scenarios, but it is possible to predict reasonable and probable events.

The scope of the project will need to be locked in at the earliest possible moment. The project team is assigned and given the task to control the well, and the project is then put into action. Usually a kick-off meeting initiates the effort.

The project management approach to the problem should include the following:

• Development of the work plan (work and organizational structures, macro project schedule, communication of the plan, and schedules establishing who does what, when, and where)

• Project planning (strategies to control the project; development of realistic schedules; responsibilities of operators, contractors, and service companies; development of a critical path model complete with network diagrams; and considerations of time, cost, and work in place)

• Design coordination (man-hour schedule, distribution of documents, drawing and procedural index, equipment index, and authority and responsibility check lists)

• Operational phase (bid packages, bid evaluations, relationships with service companies and contractors, correspondence, and reports)

• Project tracking (time schedules; cost analysis; a linking of time, cost, and work in place; and a list of who is responsible for what)

• Project completion (final inspection and assurance of completion, stopping work and demobilization, disposition of material and reusable equipment, drawings of equipment as built, and documentation of the events)

• Simultaneous tasks not directly associated with the actual kill operations (legal problems and communications with regulatory bodies, press/public relations groups, outside operators and partners, and insurers).

The project should be conducted with an early conclusion as a major goal. A key is not to be reckless by introducing additional risk at the expense of safety or by trying to reduce the cost or improve the schedule.

RESOURCES

The resources available for well control are classified as either personnel or equipment. Personnel can come from the operational staff, special contract companies, and service companies. No one will understand the local parameters better than the operational staff in place, so it is prudent to assign part of the control team from the current operational staff.

There are many schools of thought concerning who is to be in ultimate control of a blowout. Some operators believe local management is best suited for the task. Although these people are knowledgeable about the area, they may not be experienced in blowout control. (In fact, if they have performed a good job of operation they will not have blowout experience.) It is therefore advisable to use personnel who have experience with blowout and well control companies; these people should be contacted before the event occurs.

Service companies are another good source of personnel. Some service companies will be able to provide staff familiar with the types of problems encountered. They may also be able to assist by providing engineering and technical expertise in conjunction with their services.

Equipment necessary to control the well can be obtained from company-owned stores, borrowed or leased from other operators, purchased directly, or contracted from service companies. In most cases the necessary equipment is essentially common oil field equipment, such as pumps, valves, and piping.

In some cases, unique or rare equipment will be necessary. Such equipment includes pump manifolds, high pressure risers, re-entry BOP equipment, and hydraulic-set wellheads. Special, typically non-oil-field items, such as explosives and hydro-cutting tools, may be necessary. It is best if sources for each piece of equipment can be identified with pricing structures negotiated as a component of the contingency plan. Experience has shown that pricing is more reasonable if determined beforehand, rather than during the emergency.

Other special materials required may be out of the scope of normal operations. For example, a heavy brine (3.5 sg) may be necessary as a kill fluid. The sources and detailed logistics for special materials should be established beforehand.

MODELS

Some of the useful project management techniques include critical path models, decision and risk models, and resource and cost tracking models. Critical path modeling is particularly useful in scheduling and overall tracking of progress. Properly executed, the model enables the managers to predict with reasonable accuracy the cost and timing of major events. The model is also a means to force a certain amount of planning. Fig. 3 illustrates a simplified critical path model for a blowout project.

Tracking models help in reporting the cost and progress to management, partners, governments, and insurance representatives during and after the work. These models should correlate with the critical path models used.

Decision trees are also helpful in planning a complex project, especially if large groups of people need to be coordinated. Fig. 4 shows a decision tree for a typical capping operation for swamp barge operations. Decision trees force the user to think through all the contingencies that may occur during a well control problem.

ACKNOWLEDGMENT

The author would like to thank Dr. Garold D. Oberlender, professor of civil engineering at Oklahoma State University, James R. Hunt of Maxus Energy, independent drilling consultant Kent Plaster, and mechanical engineer Ralph Dean and vice-president Pat Campbell with Wild Well Control Inc. for their contributions, proof reading, and editorial comments.

Copyright 1993 Oil & Gas Journal. All Rights Reserved.