Regulatory enhancements prompted by the macondo disaster now require operators to submit a worst-case discharge (WCd) scenario report prior to obtaining a drilling permit in US offshore fields. How to model WCd for offshore wells in compliance with the new regulatory requirements is discussed here by IHS's Mofazzal Bhuiyan and Bryan McElhinney.
This article describes a workflow for building worst case discharge (WCD) scenario models for a potential blowout of an offshore well and compiling a report to comply with the new directives considering a real well from an existing Gulf of Mexico oil field. New directive (NTL No 2010-N06, effective 18 June 2010), requires a WCD report be submitted to the regulatory agency, outlining how an operator will respond to a potential blowout scenario. All offshore operators must provide a WCD report that includes the maximum flow rate, duration of flow, estimated spill volume, as well as all assumptions made pertaining to the calculation and other geological information for the well.
This article will focus on developing the required nodal analysis model, as well as gathering the necessary data to use as input for the model. While extremely important to the overall permit process, in-depth geological, log, and seismic analysis will not be covered here. Since all of these elements must be considered to develop a proper wellbore design, well planning procedures will not be covered here. In addition, considerations such as the availability of rigs for drilling relief wells and the environmental impact of a spill are outside the scope of this study.
There are several steps to follow in order to develop a WCD model that is technically sound and addresses the many concerns of potential blowout scenarios. In this study, we have explained these steps by selecting an offset producing well located in the Gulf of Mexico's South Marsh Island block 6 field. This is a deviated well with a perforation depth of 10,000ft. After modeling several WCD scenarios using a systematic workflow, this study demonstrates that this analog well would be spilling anywhere between 1128b/d and 8789b/d.
Determining reservoir parameters
Selecting appropriate analog wells to obtain geological and reservoir fluid information is one of the most important steps in the entire WCD modeling process. While this workflow does not focus on methodologies to select analog wells, great care should be taken to ensure the selected analogs closely resemble the well being modeled.
Once an analog well has been selected, the geologic and reservoir fluid parameters can be determined. Although required inputs for nodal analysis may vary, models typically used for WCD scenarios will require reservoir pressure and temperature, permeability, thickness, and areal extent. Industry software can be used to generate structure maps, do log analysis to establish producing zones (thickness), analyze faults and bounding formations to determine drainage area (areal extent), and generate the required cross-sections. In this case, IHS PETRA was used for generating structure maps and determining the thickness and extent of the reservoir. Figure 1 illustrates the South Marsh Island block 6 field formation, with well location circled in red.
Setting up nodal model
Generally, the setup for a WCD model is very similar to a regular nodal analysis for an offshore well. However, there are a few key differences: the well is expected to flow in an uncontrolled manner, pressure drops at various parts of the system would be minimal compared to a regular producing well, and non-Darcy turbulent flow is expected in almost every scenario.
In this particular case, the analog well from the South Marsh Island block 6 field is a deviated well with a perforation depth of 10,000ft that produces from the Miocene formation. In order to obtain a drilling permit to drill a new well in this formation and near the analog well, an operator would need to submit a report detailing, among other things, the expected flow rate and all relevant assumptions made to calculate said flow rate. The following steps explain the systematic process to develop a WCD nodal analysis model.
Determining the fluid properties
In case of a potential blowout, one of the most important tasks when determining a response plan is identifying the target fluid of the well. Usually, an offset well with an adjacent location or in the same producing zone can be used. In this case, the selected analog well produces primarily oil. To develop a nodal analysis model, the API oil gravity, gas specific gravity, a gas-oil (GOR) or gas-liquid ratio (GLR), and water cut need to be measured (or estimated) for this well. These will be used as the input parameters for the black-oil PVT correlations used to calculate all of the required fluid parameters, such as viscosity.
The correlations developed by Petoskey et al were specifically developed to model the fluids encountered in the Gulf of Mexico and are ideal choices for our WCD model. PVT laboratory data may be available, and can be incorporated into a WCD model to tune correlation calculated fluid properties. Table 1 shows the fluid properties available for the analog well to estimate WCD rate.
Estimating well potential
There are two main worst-case discharge scenarios that can occur. The first is when a blowout occurs during the drilling process, after a well has reached TD, but before it is completed. The second (and less likely) scenario is a blowout from a producing well with a production platform. In this case, information about the reservoirs potential to deliver fluids is relatively well known, as daily monitoring will eliminate most of the unknown parameters.
However, for a well being drilled, it can be very challenging to simulate the well deliverability. Key criteria that need to be considered for such a case are: (a) that the fluid will not follow Darcy flow behavior (non-Darcy flow is common);
(b) turbulent flow will exist in almost every situation;
(c) the near wellbore pressure drop would be limited to certain damage to the formation;
(d) multiple producing layers can contribute to the uncontrolled flow simultaneously; and
(e) adjacent producing wells may contribute to the flow.
In general, it is recommended to stick to simpler reservoir inflow models as gathering the required data for the more complex models (such as detailed wellbore and reservoir geometry) is almost impossible for new wells. For this well, the Darcy inflow model is used to simulate the well deliverability. Initial reservoir pressure has been measured at 4432psia with a temperature of 242°F.
For a well drilled on 80-acre spacing, the drainage area has a radius of 1053ft. To ensure a conservative model, we will double the inflow radius to 2106ft since contribution from nearby wells is expected. Industry software can and should be used to calculate the non-Darcy turbulence factor to account for the high flow rates encountered in WCD scenarios. Figure 2 summarizes key data pertaining the reservoir to develop potential well deliverability performance (or inflow performance) model for the WCD estimation.
Simulating wellbore flow
Properly simulating the wellbore is critical for modeling a WCD scenario. In such cases, the wellbore usually consists of casing combined with an open hole section. Typically, we assume that no tubing is present, and casing strings should be representative of the status of the wellbore. To simulate the open hole portion of a wellbore, a casing string with a diameter equal to wellbore diameter and a higher roughness factor can be used. The roughness in an open wellbore can vary widely, but is typically much higher than a casing or tubing roughness. If the known data about the well is insufficient to determine the roughness, a search on a professional organization's website may be helpful.
Usually, the effect on the total system from the open hole portion of the well is quite small.
If the well is deviated, a directional survey should be entered to account for the effects of non-vertical flow. The last step in setting up the wellbore is to assign the wellhead pressure, perforation tops and water depth. The wellhead pressure is typically examined at two values, the first being atmospheric pressure, and the second being equal to the hydrostatic pressure from the water column at that depth. Assuming that the WCD scenario occurs on the seafloor, the well will be producing with a back pressure equal to the pressure from the hydrostatic column of water. This pressure will serve to stifle the flow of the well to a degree. The other value often used is 0psig, or atmospheric pressure. This generates a true worst-case model, as the flow rates generated are as pessimistic as possible.
Current guidelines suggest that using a hydrostatic back-pressure is acceptable, although operators should be prepared to deal with spill volumes generated at atmospheric wellhead pressure as well. The perforation top depth should be set as the upper limit of the topmost perforation interval, and the water depth at the well location should be entered. Figure 3 shows the wellbore configuration without tubing to simulate an empty wellbore with casing only. In this case, the Chokshi et al mechanistic pressure drop correlation has been used to model wellbore hydraulics (outflow curve). This model was developed for prediction of flow patterns, liquid hold up and pressure drop in oil wells. Three flow patterns were considered (bubble, slug and annular) in this correlation to build the outflow curve.
Simulating what if scenarios
There are many unknown parameters when modeling WCD scenarios. It is advisable to generate several what if? scenarios before finalizing the discharge rates. For this well, two of the unknown parameters selected were average reservoir permeability and wellhead pressure, as both can vary significantly. For the first scenario, permeability is set at 100md and wellhead pressure (Pwh) is kept at 435psig. For the second scenario, permeability is increased to 200md and Pwh is set to 15psig.
The final step is to generate inflow and outflow curves to complete the calculation. In this case, IHS PERFORM has been used to run the model and generate the recommended discharge rates. Since we have analyzed two unknowns with two values, four possible WCD rates were generated. The rates will vary depending on the unknown parameters selected to perform the sensitivity analysis.
Case one yields the lowest flow rate of 1128b/d, while case four would be considered the worst-case scenario with a flow rate of 8789b/d. Figure 4 shows the well system analysis (Nodal) plot showing intersection points for the four cases. Table 2 lists all four combinations with estimated final rates, flowing bottomhole pressures and reservoir pressure drawdown values.
For this well, the reservoir drawdown pressures in Table 2 indicate that reservoir properties such as permeability are not the most important factor. The rates are dominated mainly by pressure at the wellhead, that works as aâ€™ capâ€™ in case of an uncontrolled flow situation. Hence, it is critical to focus more on modeling the wellbore, including borehole size, casing size and depth, possible wellhead back pressure and frictional pressure drops. It is also important that engineers developing the WCD scenarios perform multiple sensitivity analyses to determine the key parameters affecting the estimated WCD rates for a particular well.
A final WCD report with information about calculation assumptions and details of the modeling methods, along with final rates and pressures, then needs to be submitted to the regulatory agency in order to obtain a permit. OE