A petroleum seal is a layer of rock under pressure, against another layer of rock, along a route that may fail before anyone has counted a barrel.
Not as a figure for containment.
Not as a lid on an underground tank.
Hydrocarbons stay beneath a seal only while pore space, buoyancy, pressure, structural closure and the capillary resistance of the sealing system remain arranged in a workable geometry. Change the arrangement and the same fluids can migrate, spill, compartmentalize, mix with water, or become inaccessible to a well.
The oil can be there.
The gas can be there.
The question is whether the geometry ever held.
The Saskatchewan portion of the Williston Basin is a useful address for that question. It is not an exception to it. A drill result in any sedimentary basin can reveal hydrocarbons without uniquely revealing the route by which they arrived, the closure they occupied, or the amount still connected to the well.
That is the first difficulty in petroleum geology.
The present is evidence.
It is not the whole event.
A trap is a sequence before it is a shape
The familiar picture of an oil accumulation is a dome beneath a cap rock. Source below, reservoir above, oil caught at the top.
It is a useful drawing.
It is also a dangerous one, because it makes the trap look like a container that waited to be filled.
A working petroleum system requires source, generation, migration, reservoir, seal, trap and timing.
Each is a condition.
Their conjunction is the system.
One USGS assessment records a case where thrusting disrupted earlier stratigraphic traps and is interpreted to have permitted remigration into structural traps. The detail belongs to that assessment, not to Saskatchewan.
The mechanism travels.
A tectonic event can create a closure, open a migration route, compromise a seal, or do all three in different intervals.
Time is therefore part of the trap.
An excellent closure that arrives after the principal oil pulse may be empty. A modest older closure may be charged and preserved. A later fault can connect two volumes that were once separate, or divide a volume that once communicated.
A present-day map of structural relief cannot choose among those histories by itself.
That is not a philosophical qualification.
It is the operational reason exploration teams integrate source-rock richness and maturity, core and outcrop geochemistry, well logs, seismic stratigraphy, burial history, pressure information and oil-to-source correlation.
A regional source-rock evaluation can rest on thousands of pyrolysis and vitrinite-reflectance samples. That volume of evidence does not make a migration path visible. It shows the scale of work needed before a source interpretation becomes more than a plausible story.
One total-organic-carbon result cannot supply timing.
One oil sample cannot map the route.
One structure cannot guarantee charge.
A seal contains by refusing entry
Hydrocarbons move relative to water because they are less dense. In a communicating reservoir, gas generally occupies the highest available part of closure, oil lies below it, and water lies below oil.
The word generally is doing real work there.
Fluid contacts can be tilted by hydrodynamic flow. Capillary transition zones can broaden the change in saturation.
Faults and baffles can restrict communication.
What looks in a textbook like three clean horizontal bands can be a set of partial, locally measured boundaries.
The USGS has reported oil sourced in Upper Devonian strata occurring through thousands of feet of overlying section beneath a regional seal. The lesson is not that every accumulation travels that far. It is that a reservoir does not have to sit immediately above its source.
It has to lie on a viable migration fairway, beneath a retention barrier that worked when the fluid arrived.
This is why a seal fails by geometry before it fails by volume.
There need not be a dramatic breach. A trap can spill at its lowest closure. A fault can juxtapose reservoir against reservoir rather than against seal.
A facies boundary can disappear.
A pressure difference can reveal a baffle where seismic suggested continuity.
Containment is a relation among materials, surfaces and pressures. It is not the thickness of one impermeable blanket considered alone.
The fluid contact is a clue to that relation.
It is not a confession.
Traps ask different questions of one well
Structural traps concentrate charge where folding, faulting or salt movement creates closure and a sealing juxtaposition. Stratigraphic traps concentrate charge where reservoir quality pinches out, is truncated, or changes into a tight facies.
The two are not rival names for one thing.
In a historical USGS Malay Basin compilation, anticlines accounted for 68% of discovered recoverable reserves in the cited petroleum system. That result is local, historical, and not a global ranking of trap styles.
Its value lies elsewhere.
It shows that a basin can have a particular trap history, and that an attractive form on a seismic line is evidence only once its relevant sealing mechanism has been tested.
Fault traps require fault juxtaposition and fault-seal work. Stratigraphic traps require facies, permeability and seal-change mapping at a resolution adequate to subtle closure.
Combination traps require both.
A well placed close to a supposed boundary answers a different question from a well placed kilometres away. The distance means something only with the datum, depth and uncertainty of the mapped surface attached.
That is why a discovery well is not an answer in general form. It is an observation made at a location.
Core establishes rock properties at the sampled interval. Logs provide a continuous response down the borehole and require calibration. Seismic maps geometry between wells, at a resolution that may not capture the facies termination or fault behaviour that defines the trap. Formation pressure tests and interference tests probe communication dynamically.
Fluid samples distinguish phases and compositions.
Each data type has a support.
None of them can borrow the authority of the others.
Stored volume is not recoverable volume
A seal can hold a hydrocarbon column and still leave a difficult resource. Stored volume says pore space contains fluids. It does not say a well can draw them out at a useful rate, or that a development system can convert that rate into saleable production.
The Malay Basin assessment describes Middle Miocene sandstones with 10 to 15% porosity, and identifies compaction and authigenic clays as important causes of porosity reduction.
Porosity is a storage property.
Permeability describes the connectivity of the routes through which fluids move. Pressure support, fluid viscosity, relative permeability, natural fractures, completion design and water behaviour affect recovery again, at other scales.
The same assessment reports an average permeability near 400 mD for one reservoir set. That is useful only as an identified analogue with a stated sample basis. It cannot establish deliverability in the Saskatchewan Williston Basin, or in another interval within the same basin.
The distinction grows sharper in tight oil and shale. Horizontal placement, stimulation-created flow paths, completion intensity, well interference, production history and surface infrastructure can govern recovery from rock that contains hydrocarbons across a broad area.
Forecasting conventions carry the same warning.
A tight-oil methodology may assume a fixed well life and a decline curve that switches form at a stated annual rate.
Those are modelling choices.
They are not promises that every well will produce economically for that duration.
They make a necessary point visible.
The estimated ultimate recovery is a forecast conditioned by data and method. It is not a fluid volume found in a core.
The reservoir has not become larger because a curve has been drawn through it.
A category is not a measuring cup
Reserve language often deepens the confusion, because the words sound geological while the category is also commercial.
The Canada Energy Regulator defines reserves as remaining marketable quantities expected to be recoverable from known accumulations, using established technology, under foreseeable economic conditions. That definition joins rock, project, market and time.
It is not a count of every molecule in place.
Under the Petroleum Resources Management System, discovered quantities that are not currently commercial are contingent resources, and their low, best and high estimates are labelled 1C, 2C and 3C. Undiscovered quantities carry the analogous prospective labels.
Those classes make uncertainty and commercial status legible.
They do not make them disappear.
A field can be technically feasible and still contingent, because its market, approval, technology, development plan or access remains unresolved. Reserves can fall without oil having vanished: a changed price, schedule, facility plan, recovery performance or regulatory condition can alter the reported recoverable share.
Reserves can also rise through step-out drilling, pressure information, an approved facility or a committed drilling schedule, without the discovery of a new source rock or a larger original volume.
Canadian evaluation practice reconciles future net revenue at a 10% annual discount rate. That is a stated reporting convention for value, not a geological property of the field. SEC guidance treats development beyond five years as exceptional in its proved-undeveloped category. That is a reporting condition tied to a schedule, not a verdict on whether the hydrocarbon exists.
The quantity estimate, the development project and the reporting category have to travel together.
The route to market redraws the boundary
Now the geological geometry meets the map above ground.
Resource geography begins from an uncomfortable fact. Useful material, access capacity and benefits do not distribute evenly.
A fixed accumulation has no independent market location. It has to be reached, produced, gathered, processed, supplied, maintained and connected to a buyer, through systems with their own distances, capacities and failures.
The World Bank describes mining-corridor programmes as initially concerned with hard-infrastructure modes: road, rail and air. Petroleum systems need their own mix of roads, gathering lines, processing, power, water, communications, workforce movement and market connections.
The principle is the same.
A route built to move one product is not automatically a regional corridor because it crosses a region.
There are two linked layers in that framework. Hard assets such as road, rail, power and water. Soft infrastructure such as local enterprise capacity, skills, land administration and environmental management.
A line on a map cannot supply either layer merely by being present.
OECD evidence from Northern Ontario gives the spatial problem scale without giving Saskatchewan a coefficient. Some First Nations recorded incomes well below the regional average, and the region’s patenting rate sat below the national rural average.
Those figures do not measure an oilfield. They show why a high industrial payroll or a new export link cannot stand in for a whole regional system.
Infrastructure changes the economic boundary of the resource that justified it. A gathering system can make the next well cheaper. Processing capacity can create a market for a constrained stream. A road can serve several users, or only the original project.
Capacity, connection points, maintenance, pricing and institutional arrangements decide which result follows.
The rock stays where it was.
Its value does not.
What don’t we know?
The hardest unanswered questions are not questions of presence.
An observed oil or gas show can establish that hydrocarbons reached a location. A pressure test can show a difference between compartments. A contact can constrain part of the fluid arrangement. A seismic interpretation can supply a model of closure.
None of them uniquely identifies the migration path that charged the trap, the volume still connected across faults and baffles, or whether the observed contact reaches the full structure.
Several histories can produce a reservoir holding hydrocarbons beneath a seal. The source may have charged the trap directly, or through an earlier accumulation. A fault may have been a conduit during one episode and a barrier after reactivation. A present contact may reflect gravity segregation, hydrodynamic tilt, a capillary transition zone, compartmentalization, or an incomplete sample of a larger geometry.
The word connected carries four meanings here.
It has a geological meaning, a pressure-test meaning, a drainage meaning and an economic meaning. The first three concern paths through rock. The last concerns gathering, processing, access and market systems.
An accumulation can be connected in one sense and stranded in another.
That is the general inference problem visible in Saskatchewan’s sedimentary basin. A drill result is not a miniature history of the trap. It is a partial present-tense measurement of a system that generated, moved, retained, altered and sometimes redistributed fluids over time.
The seal is not a lid.
The reservoir is not a tank.
The resource is not a volume until geometry, recovery and the route to market hold together long enough to make it one.