Accounting for active geohazards in deepwater facilities design

An assessment combining geology, geophysics and geotechnics can aid design decisions by determining the likely effect of geohazards such as active faults and seafloor erosion on planned deepwater structures.

Michael R. Horsnell, Robert L. Little and Kerry J. Campbell, Fugro GeoConsulting

Modern offshore site investigation techniques provide data sets that enable geoscientists to combine their geophysical and geological expertise to accurately identify and characterize geohazards, both at the seafloor and over the formation depths, which will influence design decisions within the development process. During the design life of offshore facilities, identified geohazards can be categorized as either dormant—i.e., resulting from unique, relic geological events or features, with no reactivation mechanism—or active, when associated with transient phenomena or features that have a defined trigger.

Typical potentially active geohazards encountered in deepwater locations include active faults, turbidity flows, debris flows, seafloor slope failures, seafloor erosion, over-pressured soils, gas hydrates, fluid venting and mud volcanoes. In conventional foundation design, the geotechnical engineer is concerned with the influence of soil-structure interaction—i.e., how loads applied to the structure will be transferred into the soil to mobilize resistance to support the loads and to minimize associated movements. Where active geohazards are present, an additional concern will be the effect of loads imposed on the structure due to relative movement of the soil through, around and past the structure.

The integrated assessment of the geomechanical properties and behavior associated with active geohazards poses one of the greatest challenges in the design process for deepwater developments. Active geohazards can result in large-scale effects such as seafloor instability over a wide area, or microscopic effects giving rise to time-dependent changes in soil structure. In both cases, the inability to identify these problems in design could, at best, result in over-conservative design or, at worst, catastrophic failure. With respect to active geohazards, the geotechnical design process for deepwater facilities must identify the activity potential of the geohazard from current conditions and use that data to provide geotechnical characteristics for design.

ANALYZING ACTIVITY POTENTIAL

By combining the results of geophysical and geotechnical surveys, a combined seafloor model can be generated from which geohazards can be identified, Fig. 1. Dependent upon the data available and the type of geohazard, some assessment can be made of its activity history as a first step in understanding the risk of occurrence during the lifetime of the facility under design.

Fig. 1. Combined seafloor and subsurface models.

Active faults. To determine if faults are active or not and to what degree, both geophysical and geotechnical data should be used and integrated. For example, geophysical survey data allows fault characterization—i.e., mapping and determination of associated seafloor condition (seafloor scarp, burial by undisturbed strata, etc.)—indicates the character of past fault offset (sudden episodic or creep offset based on strata offset patterns with depth), and allows measurement of strata offset.

Soil samples from key strata can then be age-dated and, by making some assumptions, average rates of fault offset or offset recurrence intervals can be estimated. The reliability of fault activity prediction will vary from case to case, depending on the details of the geology and the data available.

Turbidity flows. The potential engineering significance of turbidity flows is assessed based on geophysical and geological evidence. For example, the presence of a drape of undisturbed sediment as seen on geophysical survey data indicates that turbidity currents have not occurred recently. Geophysical survey data can also help to characterize turbidity current deposits (turbidites). However, in most cases, detailed stratigraphic examination of cores is required to characterize turbidites because individual turbidites are below the resolution of geophysical survey data.

The geological evidence is then used as a framework to help estimate turbidity current speed and thickness. Age dates from cores are used to determine recency of turbidity current activity (age of any undisturbed drape) and to estimate recurrence intervals in sequences of turbidites.

Seafloor instability. Failure trigger mechanisms range from the obvious, such as spontaneous failure of rapidly deposited sediments and earthquake shaking, to less obvious and generally less common mechanisms. These include erosive under-cutting and slope over-steepening in submarine canyons and other areas subject to strong seafloor currents, slope over-steepening due to seafloor uplift caused by local salt or shale tectonics, and over-pressure in the shallow section resulting in a layer of low shear strength. Storm-wave loading does not initiate failures in deep water but may trigger failures in shallower waters that then move downslope into deep water. Certain kinds of intensified ocean currents may trigger some deepwater failures. Dissociation of gas hydrates and the resulting pore pressure buildup may also be a factor in triggering some failures, although this is yet to be conclusively demonstrated.

Assessing seafloor stability requires an integrated approach. Various types of geophysical survey data are used to characterize past failures, both recent ones with seafloor expression and ancient ones now represented by buried mass transport deposits. This allows development of a model of past failures. If soil boring data is available, dating of key strata can help calibrate the model and allow the age of the most recent failure and average recurrence intervals for repeated failure episodes to be estimated.

Seafloor erosion. Review of geophysical survey data is the first step in determining if episodic, recent or ongoing erosion and sediment transport are or have been occurring. The presence of an undisturbed sediment drape coupled with the lack of seafloor erosion/transport features (e.g., sediment waves or scour flutes) suggests that present-day seafloor currents are weak and non-erosive under natural conditions. If unburied current features are present on the seafloor, their size and orientation can provide clues as to the current magnitude and direction of flow that formed them. The presence of buried current features over a range of stratigraphic intervals as seen on geophysical survey data can indicate that strong seafloor currents have been acting for thousands of years or did so episodically over geologic time.

Sediment scour analysis based on soil geotechnical properties is performed to estimate current speeds that would be required to erode the seafloor soils. If there is no geophysical evidence for seafloor erosion, this would provide an upper limit of modern seafloor current speed. If geophysical evidence for erosion is present, the analysis would suggest that maximum currents are above the minimum calculated current speed at which erosion would occur. However, long-term (at least one year) seafloor current monitoring would be required to confirm actual flow directions and current speeds at any given site. ROV or other video observation of the details of seafloor micro-features can also provide confirmation of the degree of present-day current activity.

Over-pressured soils. As with other natural phenomena, the first step in assessing the potential engineering significance of over-pressures is to develop a geologic model to explain their origin. Interpretation of seismic reflection data provides the stratigraphic and structural framework for the model. This model then needs to be calibrated with boreholes to confirm material types and properties and in situ pressures. Once the model is calibrated, then it can be used to predict pressures away from the boreholes.

Shallow over-pressures can cause difficulty for drilling, foundation design and installation and can lead to seafloor instability and slope failures. They have been the single most expensive deepwater drilling hazard in the northern Gulf of Mexico, and have resulted in significant drilling delays and loss of many wells.

Fluid venting and mud volcanoes. The key to assessing the significance of pockmarks and mud volcanoes is to determine the degree of ongoing and likely future activity. Assessment begins with detailed mapping and characterization of both seafloor and buried features based on interpretation of high-resolution seismic reflection data. Sizes, shapes, stratigraphic (temporal) and positional evolution, extent of sediment expulsion aprons, and other details are important to developing a valid geologic model. Age dating of “normal” sediments sandwiched between expulsion deposits can sometimes be undertaken to determine recency of activity or to estimate expulsion recurrence intervals, and thereby “calibrate” the model. Direct observation of pockmarks and mud volcanoes using ROVs can sometimes help to confirm recency and degree of expulsion activity.

The model can then be used to help develop reasonable standoff distances. Because details are important to assessing these features but data collection requirements (both time and costs) may be prohibitive in areas with numerous expulsion features, “model” studies are sometimes performed. In these, a limited number of representative features are studied in some detail (e.g., using very closely spaced survey lines by autonomous underwater vehicles, or AUVs) as a model for characterizing other similar features in the area.

Gas hydrates. There are three mechanisms by which gas hydrates form in deepwater environments:

1) Low-concentration, widespread accumulations of methane hydrates (mostly of shallow biogenic origin)

2) Localized high-concentration deposits of heavier, petrogenic hydrates (formed from gas that has leaked upward into the stability zone from reservoir depths along pre-existing faults and around vent features)

3) Medium-concentration combinations of the two types.

This last type of hydrate formation results from migration of pore fluids containing petrogenic hydrocarbon gases in solution from reservoir depth into the shallow section.

Widespread methane hydrate accumulations are sometimes accompanied by a bottom-simulating reflector (BSR) on seismic reflection data. The BSR can represent the accumulation of small amounts of free gas trapped at the base of the methane hydrate stability zone. In some areas, a second and often more localized BSR is seen below the regional methane hydrate BSR. This deeper BSR can be interpreted to represent the base of the stability zone for the petrogenic hydrates that result from upward-leaking heavier hydrocarbons.

Seismic data alone are not adequate to identify and map the distribution or concentration of hydrates. To date, a combination of downhole geophysical logging, in situ testing, pressure coring and laboratory testing of host sediments has been the only reliable methodology for hydrate identification and characterization.

DESIGN CONSIDERATIONS

Dependent on the type of geohazard and its activity potential, it may be advisable to avoid the area of seafloor concerned altogether. However, if activity potential is considered marginal or there are economic benefits to designing around the influence of the geohazard, then analyses need to be undertaken to understand the present-day condition, to provide a springboard to the facilities design. Techniques available will depend upon the geohazard under consideration. Having a method for analyzing the geotechnics of a potentially active geohazard requires that one also have a reliable set of parameters to use within that analysis. Most of the analytical techniques outlined in Table 1 have a proven track record for more traditional soil conditions encountered on the continental shelf (e.g., clays, silica sands), but their application to soils encountered in deep water and their geological setting may need to be calibrated.

TABLE 1.  Methods appropriate for analyzing geohazards in deep water dependent upon geological setting

Active faults. Considering potential damage to wells and the potential effects on foundation reliability, development plans are typically designed to avoid active faults. In many instances, however, significant benefits to the development plan can be realized by knowing how close foundations or pipelines can be safely placed to these faults. This is normally achieved by establishing an exclusion zone around the fault, Fig. 2.¹

Fig. 2. Cross-section showing exclusion zone relative to fault and anchor pile.

The width of this zone is defined by the uncertainty in the seafloor location of the fault, the dip angle of the fault, the penetration depth of the foundation element, the positioning uncertainty in locating the foundation element, and the extent of soil that will be relied upon in the foundation design (the foundation load influence zone). Numerical or analytical modeling can be preformed to determine the foundation load influence zone.

In some locations, the cost of routing pipelines around existing faults may also be prohibitive or result in flow assurance difficulties. In these cases, integrated modeling of fault displacements with soil-pipeline interaction studies may be used to determine the stresses induced in the pipeline as a result of fault movement. Results from these studies may allow the pipeline to be designed to tolerate the anticipated stress.

Turbidity flows. Turbidity currents may achieve velocities sufficient to displace and damage surface-laid pipelines and umbilicals and induce scour around foundation elements. Quantitative engineering analyses can be undertaken to evaluate the potential effects of turbidity currents on the development facilities after establishing the geologic model for past events. Details gathered in the field investigation are used as input, including the seabed morphology, the distribution, grain-size character and thickness of deposits from past turbidity flow events, erosivity and settling velocity of the source sediments, and probable triggering events (seabed instability, canyon flows from off the shelf, etc.). These inputs are used, through parametric studies, to calibrate quantitative numerical models in order to predict the character of potential future turbidity flows.²

These models can provide the engineer with predictions of the average velocity and density of flows in time and space, which can, in turn, be used to derive loads imposed on surface-laid pipelines and structures.^3–5 This information allows the designer to consider mitigation schemes, such as increasing pipeline weight, trenching and sheltering of facilities.

Debris flows. Debris flows have the potential to reshape the seabed. As the flows travel downslope, they can mobilize the underlying seabed sediments through shear transfer, they can overload the seabed and cause bearing failures, they can erode and entrain seabed sediments along the flow path, and they can leave thick deposits along the length of their route and at their termination. Buried deposits may consist of materials with highly variable geotechnical properties, including rafted blocks of much stronger, intact strata compared to the enclosing matrix.

Such heterogeneity can complicate soil characterization and, ultimately, foundation design and installation plans. Active flows can impose prohibitive loads on seabed structures and pipelines. Often, locating facilities outside the path of potential debris flows is the preferred design solution.

With a detailed geologic model of past debris flow events and knowledge of the geotechnical properties, quantitative debris flow modeling can be preformed to predict probable flow paths of future events. As with the turbidity currents, the diagnostic models are performed on past events to determine the most appropriate flow parameters. Prognostic modeling of future events then provides estimates of flow paths and flow characteristics (such as thickness, velocity and acceleration), and how they vary in time and location. This information can then be employed to estimate potential loads induced by the flow on seabed facilities.^6,7 However, designers should be aware of the limitations of the models, particularly in their ability to capture all aspects of the potential effects of the debris flows on the seabed. To fully account for the potentially hazardous effects of future debris flows, the geologic record should be carefully investigated.

Seafloor slope failures. Seafloor slope failures can result in prohibitive loads on seabed facilities located within the failure zone, can be the source for debris flows and turbidity currents affecting facilities downslope, and can destabilize upslope areas. The most obvious hazards presented by these failures, depending on facility location with respect to the failure, include catastrophic lateral loading and loss of support that can be applied to structures during failure events. Mitigation measures may be achievable for smaller-scale failures, but often avoidance of potential failure areas is the preferred solution.

Quantitative slope stability studies can be approached in a variety of ways, including equilibrium-based 1D or 2D analytical modeling, 2D or 3D finite element modeling, total stress or effective stress parameters, and static, pseudostatic or dynamic loading. Qualitative geologic assessment of past slope failures and their recurrence intervals provides a basis for selection of the most appropriate analyses and offers the engineer the opportunity to investigate how well the selected engineering parameters and methods model past events.

When the geologic assessment and diagnostic geotechnical analyses are mutually supportive, the results of prognostic modeling will have greater reliability. The quantitative results provide a basis for evaluating the probability of future slope instability and the impact area influenced by events of different scale. This information can be used to determine safe standoff distances or to derive parameters for input into models for evaluation of effects for mitigation designs.

Seafloor erosion. Seafloor erosion and sediment scouring can result in the loss of support around foundation members and beneath pipelines. Conversely, in areas of high sediment transport, large, unexpected amounts of sediment can build up around or bury facilities, or can quickly fill excavated pipeline trenches before pipe laying is carried out. Local scour around seabed structures is usually the result of the intensified vortices in the current as it flows across the foundation elements. Local scour degrades foundation performance.

Numerical simulations can be constructed to model the shear stress induced in the seabed sediments around the foundation elements under an applied current. Current parameters may be supplied from metocean studies, for long-term currents, or possibly from turbidity current modeling results. The scour rate can then be evaluated based on sample test results in an erosion function apparatus.⁸ The effects of the predicted scour can then be examined analytically to determine what level of mitigation may need to be considered.

Gas hydrates. Although the potential for naturally occurring gas hydrates to cause seabed and sub-seabed deformations and stability problems for wells and seabed-founded structures has long been a concern, the only reported case histories we are aware of are limited to Arctic permafrost regions with sand and gravel hydrate-bearing layers. Little has been published in terms of detailing the risks to facilities in hydrate-bearing sediments; however, to quantitatively assess the risks requires analytical estimates of the effects of gas hydrate dissociation on the soils and structures.

Analytical assessment of the effects of producing wells passing through naturally occurring gas hydrate zones is difficult, as many variables affect gas hydrate dissociation. These variables include:

1) The geochemistry of gas hydrate present in the soil column, the physical form of the gas hydrates, their distribution, and their volumetric percentage

2) The physical characteristics of the soils overlying, within and vertically adjacent to the gas hydrate-bearing zones, the thermal properties of the soil, and the in situ state of stress in the soil

3) The design and installation details of the wells, thermo-conductivity of the well components, and thermal characteristics of the product pumped through the wells.

Given the complexity of the problem, lack of an empirical database and the potential range of responses depending upon the variables above, exacting analytical methods may be in excess of what is required or suitable at this time. However, the effects of gas hydrate dissociation related to hydrocarbon production can be estimated using simplified geologic/geochemical models of the gas hydrate and its distribution and approximate numerical models of heat transfer, gas hydrate dissociation, fracture propagation and seabed deformation.

The results of these models can be used to estimate the areal extent of seabed where warming of the soil will lead to hydrate dissociation, the volume of gas expected to be released due to hydrate dissociation, the stability of proposed production wells subjected to subsurface sediment movements and induced stress through load transfer from the soil to the well casing elements, the potential influence on seafloor foundations within the vicinity of the production wells, and the potential relative vertical and horizontal seafloor surface motions within the vicinity of the production wells, Fig. 3. With the estimated soil response and the effects explored in these models, risks to foundations and wells can be better understood and quantified, and mitigation measures can be evaluated.

Fig. 3. Gas hydrate dissociation and soil response: Conceptual models for volumetric seabed strains.

Over-pressured soils. Over-pressured soils may pose a drilling hazard, can result in unconservative foundation design, and lead to an underestimation of geologic hazards. In terms of over-pressured soils as a drilling hazard, shallow water flow in a well poses one of the greatest concerns to drillers. Mitigation is usually based on 1) avoidance of zones of shallow water flow-prone sand layers identified in the geologic model, 2) well and casing design, and/or 3) precautionary drilling practices.

When shallow water flow occurs, accumulated sands around the wells can make placement of seabed facilities difficult, forcing changes to development plans. Over-pressured soils in the shallow section can also be a factor in seabed stability evaluations and in foundation design. Most often, offshore foundation design and evaluation methods are based on the assumption that the in situ pore pressure is equal to the hydrostatic pressure. Formations with pore pressure in excess of hydrostatic pressure, however, will have lower soil strengths than similar formations with no over-pressure. While analyses can be performed using effective stress methods, direct in situ measurements to determine the pore pressure are not commonly obtained. If unidentified over-pressured formations exist, foundation capacities and response could be overestimated and the threat posed by geohazards, such as slope instability, could be underestimated.

It is recommended that geotechnical laboratory and in situ testing programs be designed to assist in the identification of over-pressured soils and that the geologic model be consulted during the selection of engineering parameters for foundation design or geohazard modeling.

Fluid venting and mud volcanoes. Avoidance is the recommended mitigation where active vent features are encountered. Safe standoff distances should be based on an evaluation of the degree of activity, the areal extent of surface effects, and the subsurface conditions. Soil conditions around extinct or dormant features should be determined if foundations or pipelines are to be placed nearby in order to ascertain the soil strength and stability of the sidewall slopes. High fluid intrusion in the soil and mud flows during periods of active expulsion could have had long-lasting effects on the soil, making local conditions much different from those at sites located well away from the features.

CONCLUSIONS

Despite the challenges of active geohazards, an assessment combining geology, geophysics and geotechnics can be made of the effect that any identified geohazard will have on the design of the facility under consideration. This may confirm that the geohazard can be incorporated within the design or has to be avoided altogether.

ACKNOWLEDGMENT

This article was prepared from “The geotechnical challenges of active geohazards in the design of deepwater facilities,” presented at the Society for Underwater Technology Annual Subsea Technical Conference held in Perth, Australia, Feb. 17–19, 2010.

LITERATURE CITED

 ¹ Campbell, K. J., Burrell, R., Kucera, M. S. and J. Audibert, “Defining fault exclusion zones at proposed suction-anchor sites using an AUV micro 3D seismic survey,” OTC 17669 presented at the Offshore Technology Conference, Houston, May 2–5, 2005.
 ² Reed, C. W. et al., “Analysis of deepwater debris flow, mud flows and turbidity currents for speed and recurrence rates,” presented at the Deepwater Pipeline and Riser Technology Conference, Houston, 2000.
 ³ Chakrabarti, S. K., “Loads and responses,” in Chakrabarti, S. K., ed., Handbook of Offshore Engineering, Elsevier Science and Technology Books/Harcourt, London, 2005.
 ⁴ Dalton, C., Reed, C. W. and A. W. Niedoroda, “Determination of forces due to a current on a pipeline in the vicinity of a seabed,” paper 5044 presented at the 19th International OMAE Conference, New Orleans, Feb. 14–17, 2000.
 ⁵ Det Norske Veritas, “Environmental conditions and environmental loads,” Recommended Practice DNV RP C205, Hovik, Norway, April 2007.
 ⁶ Marti, J., Lateral Loads Exerted on Offshore Piles by Subbottom Movements, PhD dissertation, Department of Civil Engineering, Texas A&M University, Report No. MM 3008-76-5, 1976.
 ⁷ Schapery, R. A. and W. A. Dunlap, “Prediction of storm-induced sea bottom movements and platform forces,” OTC 3259 presented at OTC, Houston, May 8–11, 1978.
 ⁸ Briaud, J.-L. et al., “SRICOS: Prediction of scour rate in cohesive soils at bridge piers,” Journal of Geotechnical and Geoenvironmental Engineering, 125, No. 4, April 1999, pp. 237–246.
 

THE AUTHORS
	Mike Horsnell has over 30 years of experience in offshore geotechnical design, including pile design and pile group behavior, the response of offshore foundations to cyclic and dynamic loading, and pipeline-soil interaction. In 1983, he was appointed Manager of Engineering of Fugro Intl., where he gained experience in offshore soils in the Gulf of Mexico, offshore California and Alaska. He was appointed Managing Director of Fugro Ltd. in 2005, and subsequently of Fugro GeoConsulting Ltd. In 2008, he was appointed President of Fugro GeoConsulting Inc., where he also holds the responsibility of Fugro’s Worldwide Director of GeoConsulting.
	Rob Little has 21 years of experience in the offshore geotechnical field. As Senior Consultant with Fugro Geoconsulting, he has been responsible for directing geotechnical design projects, integrated site characterizations and geohazards assessments. These projects have included the analysis of gravity-based structures, tension leg platforms, underwater slope instability, wave-seabed interaction, pile driving in difficult soil conditions, production-induced subsidence and cyclic degradation under both wave and earthquake loading.
	Kerry J. Campbell began his career in 1968 as a Research Geologist with the US Geological Survey. Subsequently, he carried out sea-ice and permafrost surveys in the Arctic using radar profilers followed by site assessments for coal mine development in West Virginia. For the past 34 years, he has been a Marine Engineering Geoscientist with Fugro and predecessor companies. Mr. Campbell has been responsible for consulting, project supervision, field mapping and data acquisition, data analysis, technical reporting, and business development for various types of offshore and onshore geologic and geophysical projects worldwide. He is currently Principal Geoscientist in the Houston office of Fugro GeoConsulting.