Abstract

Operators are looking at drilling for deeper gas reservoirs. There are a number of problems unique to ultra-deep drilling - greater than 20,000 feet. Well control is complicated by narrow margins and higher bottom hole pressures and temperatures. Application of tools that can assist in pore pressure prediction are limited by the depth, temperature and the high cost of errors. Well control problems can be spectacular. The depth pushes rig equipment loads to their design ratings. Maintenance, inspection and design of all components are more critical due to reduced margins. Wellbore tubulars are often designed for sour and corrosive production fluids. Those designs increase the pipe's weight and cause special handling requirements. Drilling tools are also limited by the depth, temperature and rock hardness. Two operators working in Texas, Louisiana and Wyoming drilling to depths of up to 25,000 feet have documented many of the industry's problems and current solutions. Understanding the problems and designing wells to reduce the risks is the first step to a cost effective drilling operation.

Introduction

This paper considers high temperature to be greater than 300° F, high pressure to require equipment rated to more than 10,000 psi, and ultra-deep to be in excess of 20,000'. The experience base is greater than 400° F, 18,000 psi and 23,000'.

Interest is increasing in deep drilling. In the Gulf of Mexico (where data is readily available) the MMS estimates 5 to 20 TCF is in deep reservoirs1. Only 5% of the 35,000 wells drilled in the Gulf have been drilled deeper than 15,000'. The MMS estimates the average size of 500 discoveries from below 15,000' to be 20 BCF. The MMS has offered royalty relief on the first 20 BCF produced from deep wells. On land the deep reservoirs have not been as thoroughly drilled as the shallow basins. There is similar potential for larger reserves recovery at higher flow rates and thus, greater return on the investment. All trends indicate interest will continue to grow.

Deep drilling has problems. Costs are high. Mechanical and exploratory risks are high. Smith2 documented an exponential relationship between drilling costs and depth based on API data: Cost = 80000 e 0.000255*Depth

The problem of high costs is best "solved" by eliminating the "Train Wrecks." Minimize the risks inherent in deep hot wells with detailed plans and thorough communication. Most individual problems can be addressed and "solved" during the design and planning phase. Steps can be taken to minimize risks and the potential for many execution phase problems. The challenge in the design phase is addressing the problems and building flexibility into the well plan. Problems and plan changes can result in a well not achieving its objective when contingency casing strings cannot be included in the preliminary plan. Often the realization that the objectives will not be met occurs after a considerable investment. Drilling incentives on royalty relief can only pay out if discoveries are made.

This paper has separated the problems of ultra-deep wells into four primary topics. The best time to address and avoid problems is during the design and planning of the well. The focus at pre-spud should be toward risk mitigation; identifying potential problems and taking steps to avoid them. Well control is the most critical problem that must be addressed by the design. A major factor in the cost is time on location due primarily to slower rates of penetration deep in the wells. These problems are addressed in drilling issues. Finally, completion problems must be addressed to provide for production.

Risk/Trouble Mitigation

Risk/trouble mitigation for onshore ultra-deep wells is addressed during the design, planning, drilling and operations as it is any well. However, these wells require a higher degree of investigation, conceptualizing during the planning and design and communication with a larger team and for longer periods of time.

This content is only available via PDF.
You can access this article if you purchase or spend a download.