Keeping the hydrocarbons in the trap, effective seals in petroleum exploration

Why are seals so important in petroleum exploration?

There are five elements in any petroleum field, all of which have to work if the field were to contain hydrocarbons. If any of these elements fails then we will have a dry hole. The elements are:

a.      A source rock which is normally an organic rich shale, the organic matter is transformed into hydrocarbons by heat and pressure. The source rock is mature when it is generating hydrocarbons and super mature when the hydrocarbon generating potential has been exhausted

b.     A reservoir rock. This is a porous rock, normally a sandstone of limestone , which holds the hydrocarbons

c.      A trap. Hydrocarbons are buoyant and will float up to the surface if they are not trapped. Taps can be structural such as anticlines or fault blocks or stratigraphic such as pinch-outs.

d.     Migration. Hydrocarbons have to migrate from the source rock kitchen to traps containing reservoirs. This has to be done at the right time, for instance after the traps have been formed by tectonic processes.

e.     Seals. This is a cap rock that ensures that the hydrocarbons stay in the trap. The Cap rock has to hold back the pressure of the buoyant hydrocarbons and not be damaged by tectonic movements. Stratigraphic traps which are formed by lateral changes in rock type also require side seals and bottom seals as well as top seals to prevent hydrocarbons escaping

Seal effectiveness is the hardest of these elements to predict. A great deal of effective technical work is done on reservoir prediction using sequence stratigraphy, sedimentology and reservoir geology. Basin modelling or petroleum systems modelling (and geochemistry) looks at the source and migration petroleum system and can help to reduce the risk on hydrocarbon charge and help predict the hydrocarbon phase. Seismic interpretation and structural geology looks at defining the trap. Seismic attribute, AVO, inversion and other quantitative interpretation (QI) work can help with reservoir prediction and also with predicting charge,

Seal prediction however has been a bit of a Cinderella in petroleum geology. Many geologists would almost take seals for granted, particularly in a shale rich system or would overestimate seal risk.

This article is influenced by John Karlo’s lecture “your next dry hole is most likely to be caused by seal failure” which can be seen here - https://www.youtube.com/watch?v=F_YyEpfoj0s

How likely is Seal Failure?

Seal Failure is the most likely cause of dry holes in mature plays. In the early stages of exploration, for example in frontier basins, reservoir distribution, source rock distribution and maturity are poorly understood. Explorers are testing new ideas which most of the time will not work (success rates in frontier basins are about 10%). Eventually after some tears, tantrums and dry holes, the geologists make some discoveries and begin to work out how the petroleum system functions and where the good reservoirs are. Success rates increase to about 35% to 40%, but this means that the majority of exploration wells still fail. At this stage seal failure becomes the most likely cause, particularly as more complex stratigraphic traps are drilled.

Figure 1. Well Failure analysis by cause.

Figure 1 shows two analyses of well failure. The left hand chart shows an analysis done by Schlumberger which shows that 45% of the wells that they looked at failed due to seal (14 % due to top seal and 31 % due to lateral seal). The right hand chart was compiled by the UK oil and gas authority. This is part of a comprehensive report on post well analysis in the Central North Sea basins, of wells drilled between 2003 and 2015. This report is available here (https://www.ogauthority.co.uk/media/1578/cns-mf_post_well_analysis_report.pdf ). The Central North Sea (including the Moray Firth) is a mature are which is fully covered by 3D data. Many of wells drilled in the last 15 or so years have a stratigraphic trapping component and lateral seal (27.3 %) and bottom seal (5.4%) are the main culprits.

Figure 2. Well failure cause by play maturity.

Rudolph and Goulding (AAPG Bulletin, Feb 2017) have looked at benchmarking exploration wells drilled by ExxonMobil over the last 20 years. They subdivided the wells drilled by play maturity, 54 wells were drilled in new plays, 38 wells extended discovered plays into new areas and 138 wells were drilled in mature areas. Petroleum charge (source and migration) was the key risk element in frontier plays, Seal and charge in extending areas and seal was the overwhelming key risk element in mature areas.

How do seals fail?

J. Karlo stated in his lecture that there are five main causes of seal failure (with my comments):

There are three main causes for top seals

1. The micro-porosity within the seal enables hydrocarbons to flow upwards when the buoyancy of the hydrocarbons exceeds the capillary entry pressure in the seal rock. This is dependent on pore size within the shale (clay content, compaction, cementation etc.) and hydrocarbon type. This is also a significant cause for side or bottom seal failure.

2. A brittle top seal is fractured by faulting, enabling hydrocarbons to escape via the fractures. This may also apply to some side and bottom seals

3. Seal failure due to hydraulic fracturing by high aquifer and hydrocarbon column pressure which exceeds the fracture pressure of the seal rock.

There are two main causes for fault related seals

4. Fault displacement causes the failure of shale rocks smeared within the fault

5. Fault rocks contain silt and sand which enables cross fault flow, leading to trap leakage.

Figure 3 – Potential seal failure points

The diagram in Figure 3 shows areas for potential seal failure. The dark blue line shows the contact if the top seal only is intact (i.e. the fault seal fails), the mid blue line shows the contact in the case Fault seal failure in the sand on sand contact zone and the light blue line shows potential contact in case of successful sand on sand fault seal.

Types of Seal rocks, some shales are more equal than others.

There are three main types of cap rock:

1.      Evaporites, such as halite. Rock salt that had been deposited in ancient dried out lakes or enclosed seas, think Bonneville Salt Flats in larger magnitudes. This type of rock has no permeability and should form a perfect seal. However there may be gaps in the salt due to salt related tectonics or other leak paths formed by rafts of porous dolomite so a salt top seal is not a given

2.      Tight limestones and chalks. These are limestones that have very little to no porosity left and have been heavily cemented. Chalks form top seals for several Triassic and Jurassic North Sea fields (e.g. Jade and Judy) and bottom seals for several North Sea Palaeocene fields with stratigraphic traps, such as Everest and Fleming. However there is a risk that there may be some remnant porosity which may cause a leak and limestones are more brittle and prone to fracture failure than shales.

3.      Shales and other mud rocks. These fine grained siliciclastic rocks have very low porosities and small pore throat sizes that require a higher capillary entry pressure to break through. Shales are also more ductile and less prone to brittle failure. Shale seal potential is also heavily dependent on facies. This is described by Dawson, Almon , Dempster and Sutton (2008) – their paper can be accessed here - https://www.searchanddiscovery.com/documents/2008/08144dawson/ndx_dawson

Dawson et al identified the following shale facies

1.       Type 1 - well laminated, high clay content and high organic content, deposited in deep water - best seal capacity (8395 psia 10% mercury injection capillary pressure (MICP), a measure of seal capacity)

2.      Type 2 – some lamination. High clay content but lower than type 1, more silt content than type 1, as well as more carbonate, high organic content – next best seal capacity (7445 psia MICP)

3.      Type 3 – more silt and carbonate than type 2, lower organic content, mottled texture, deposited in slope settings – less good seal capacity (4950 psia MICP)

4.      Type 4 – more silt and carbonate than type 3, lower organic content, mottled silty texture, deposited in slope settings – less good seal capacity (3175 psia MICP)

5.      Type 5 – more silt and carbonate than type 4, lower organic content, laminated silty texture deposited in upper slope settings – lowest seal capacity (1360 psia MICP)

6.      Type 6 – A shale that has had enhanced cementation reducing porosity and permeability, massive with few sedimentary structures (7665 psia MICP)

Type 1, 2 and 6 facies, laminate or cemented shales formed the best seals, while slope mudstones formed poorer seals. Organic content was related to increased seal capacity, while silt and carbonate content was related to poorer seal capacity.

A short introduction to pressures

Pressure within the fluids contained in the rocks that they contain increases with depth due to the increased thickness of overburden rocks according to the formula Rho * g * h = pressure, where rho is the fluid density, g is the gravitational constant and h is the height of the column of fluid. A few definitions:

·        Pore pressure is the pressure of the fluid contained in the pores within the rock, caused by the weight of the fluid column only. In normally pressured rocks this pressure lies on the hydrostatic gradient (ranging between 0.433 psi /foot (9.792 kpa/m) and 0.465 psi /foot (10.516 kpa/m) depending on fluid density)

·        The lithostatic gradient is the pressure exerted by the weight of the rock and fluids in the overburden. The gradient varies depending on rock type but is normally about 1 psi/foot or 23 Kpa/m.

·        Fracture pressure (or Frack pressure) is the pressure at which the rock would begin to break down and form fractures. If the pore pressure exceeds the rock’s frack pressure then fractures develop allowing the fluids to leak. If the fluids are hydrocarbons then they will leak upwards breaching the trap. Fracture pressure gradients are typically about 0.7 psi / foot (15.8 Kpa/m) depending on lithology.

·        Overpressure in pore pressures is when the pressure within the pores is higher than that predicted by the hydrostatic gradient. This can be due to the aquifer/reservoir being isolated during rapid burial.

·        Hydrocarbon pressures. As hydrocarbons, particularly gas are less dense then water, they have steeper pressure gradients, typically 0.08 psi/foot for gas and 0.35 psi/ foot for oil.

·        Datums. On land all pressures are measured relative to ground level. Offshore fluid pressures would be measured from sea level (or lake level), but rock pressures are measured from seabed. This means that the datums can be at significantly different levels when drilling in deep water with a much narrower gap between the lithostatic and hydrostatic pressure levels leading to weaker seals.

·        Capillary entry pressure is a measure of the inherent sealing capacity of a rock. This is the pressure needed to enable a non-mixable fluid (such as hydrocarbons) to displace another fluid such as water within the pores of a rock by overcoming surface tension. The CEP depends on porosity, pore size and shape as well as exact fluid properties (wettability). Smaller and more tortuous pores lead to a higher CEP and hence can hold back a greater hydrocarbon column. Capillary entry pressures are measured on rock samples in a laboratory by injecting mercury. For further reading please see reference = https://store-assets.aapg.org/documents/previews/1044ST60/CHAPTER01.pdf . The most common cause of seal failure is when the pressure exerted by the buoyant hydrocarbon exceeds the capillary entry pressure of the sealing rock allowing the hydrocarbons to percolate upwards.

Figure 4. Diagram showing pressure definitions.

There is also a useful video explain pressures here - https://www.youtube.com/watch?v=k2TufD6w0JU

Pore pressures are measured directly by pressure logging tools such as RFT and MDT. Pore pressures can be inferred via drilling mud weights. Pore pressures may also be predicted using seismic velocities (please see here for an explanatory video - https://www.youtube.com/watch?v=sUhtFJL6_0k ). Fracture pressures may be measured by leak off tests or formation integrity tests. All of these methods, except seismic requires prior drilling in the local area. Pressures within an undrilled prospect can be modelled by extrapolating data from adjacent wells. There is obviously less accuracy in predicting using offset wells which are far away in a less mature petroleum province.

Fault Related Seals.

Fault sealing is not that well understood. Some faults are seen as conduits for hydrocarbons, enabling migration from a deeper kitchen to a shallower reservoir. Alternatively a fault can be the key component of a trap, forming a seal between compartments within a trap. Occasionally I have seen the same faults acting as both conduit and seal as it would suit the promoter of the prospect. There are two basic type of fault seal (as shown in Figure 3 above):

a.      Sand on shale contact. This is when a reservoir sandstone is juxtaposed with a shale across the fault. The shale is believed to be intrinsically sealing, but leakage is possible up the fault plane or via fractures associated with the fault.

b.     Sand on Sand contact. This is when the reservoir sand is juxtaposed against porous and permeable sand across the fault. In this case the fault rocks, including gouged up shales or clay smears in the fault zone, might or might not form a seal retaining the hydrocarbons in the trap.

It is also common for the fault zones in sand on sand contact to have some sealing potential, even if they cannot hold back the entirety of the hydrocarbon column forming a complete seal. This leads to different contacts in separate fault blocks within faulted field. Faults may also form a barrier or baffle during production and may also have differential leakage, letting through gas but holding back oil.

A good but long video on fault seal can be found here - https://www.youtube.com/watch?v=iU5B73DechU

But I have a Seismic Anomaly – surely the trap must be intact?

Well, err, not really. A low saturation gas sand, with saturations as low as 5% can still give a seismic direct hydrocarbon indicators (DHI) including potentially AVO, flat spots and velocity push downs.

This video by Dr Mijke Van de Boogard looking at shallow gas deposits in the Netherlands includes modelling of both high and low saturation gas sands https://www.youtube.com/watch?v=8EAXn-o9C7k

In a play where hydrocarbon fields normally has a seismic expression, a trap that has leaked but still has residual gas is likely to have similar potential DHI’s to an intact trap

Summary

·        Seals are the hardest of the elements of petroleum geology to predict, but some methods exist both qualitative and by statistical analysis of nearby discoveries ( column height analysis)

·        Seal failure is the most common cause of dry holes in mature basins. Seal failure is also a major dry hole cause in emerging and frontier basins

·        Seal analysis is a bit of a Cinderella in prospect analysis. Most of the time is spent in seismic interpretation focussing on defining the trap, sedimentology defining the reservoir and perhaps basin modelling defining the source and migration. Some geologists can take sealing for granted or alternatively say that it is too complex to model effectively.

·        There are few geophysical methods that are potential “silver bullets” that can show that a trap is breached. One example of a seismic tool for trying to analyse leakage is the de Groot Bril chimney cube. This technique is explained in the Leading Edge Journal, May 2001 issue. A video on using this tool is available here - https://www.youtube.com/watch?v=Eh76P_VSyEw . 

However having a seismically visible gas chimney does not necessarily mean that that trap is breached, as lighter fractions only (gas) could have leaked leaving oil behind or the trap is in dynamic equilibrium with a source charging faster than leakage. This subject is not well understood.