🛢️ Oil and Gas: How Wells Find and Free Hydrocarbons
Follow the industry upstream: seismic surveys that image rock miles down, drill bits that steer, and the completion work that makes a well produce. You'll understand reservoirs, fracking, and offshore
What you’ll learn
- There Is No LakeReplace the cavern mental model with rock, separate porosity from permeability, and adopt the course's through-line: upstream money buys information or permeability.There is no underground lake — oil sits in the pore space of solid rock, like cooking oil in a brick, so the question is never 'where is the pool' but 'which rock, and will it yield'. Porosity says how much oil is present; permeability, measured in darcys (about 10⁻¹² m², a unit of area) and in practice millidarcys, says whether the pores connect into paths that let it out, and a rock needs roughly 100 md to be exploitable without stimulation. The cruelty is that the same property below that threshold makes an efficient seal, so every field is a collaboration between permeable reservoir and impermeable cap — an invisible arrangement kilometres down, which is why every dollar upstream buys either information about unseen rock or permeability nature withheld.
- What You Are Actually HuntingSpecify the petroleum system as a four-part coincidence — source, migration, reservoir, seal and trap — and recognise the trap geometry as the object seismic imaging is hunting.Oil forms where organic-rich mud was buried in anoxic water and cooked in the 60–120 °C oil window, almost always a shale; it is then buoyant and migrates upward for millions of years, which is why the tar pits and seeps humans have always known are failed fields. Accumulation requires a permeable reservoir rock, an impermeable seal, and — the rarest item — a closing geometry, since a seal alone merely diverts migrating oil sideways. Anticlines trap oil under an arch with gas, oil and water sorted by density; fault traps butt permeable rock against impermeable; salt domes rise and bend strata into traps while wrecking the seismic image; and stratigraphic traps, where a sandbar or reef simply pinches out, have no structure to see at all.
- Seeing With SoundUnderstand reflection seismology as the industry's central skill, and derive from acoustic impedance why seismic images contrasts rather than rock.A trap is a few hundred metres across, kilometres down, with no surface expression, so the only option is to shout at the ground and listen — reflection seismology, on which every later decision depends. Echoes occur at contrasts in acoustic impedance Z = vρ, with reflection coefficient R = (Z₂ − Z₁)/(Z₂ + Z₁): matched impedances give no echo however different the rocks are geologically, and a drop in impedance flips the reflection's polarity, which interpreters read. Sources are dynamite (a near-perfect impulse) or vibroseis (a swept chirp, later compressed computationally) on land and air-gun arrays at around 2,000 psi and 2,000–7,000 cubic inches at sea; receivers are geophones and hydrophones; and the measurement is two-way travel time, t = 2d/V — an equation that appears to solve the problem until you notice V is unknown.
- From Echo to ImageExplain CMP stacking and migration as the processes that build an image from noise, and treat the seismic image as an inference with known failure modes.A single reflection is buried in noise, so every subsurface point is sampled by many source–receiver pairs sharing a midpoint; correcting each trace for its path length and summing the common midpoint gather makes the coherent signal grow while random noise cancels, which 'significantly lowers the random noise'. Migration then relocates each event from the surface position that recorded it to its true subsurface position — necessary because tilted reflectors echo from off to the side, smearing exactly the steep structures that traps are made of — but it requires a velocity model that is itself unknown, so imaging proceeds iteratively on very large computers. 3D seismic, first acquired at scale in the late 1970s and widely used through the 1980s and 1990s after 1960s experiments, changed what could be sought by revealing stratigraphic traps with no structure; bright spots remain a real but treacherous direct hydrocarbon indicator, since 'fizz gas' at a few percent saturation can look nearly as bright as a commercial accumulation.
- The Well Is a QuestionUnderstand a well as a pressure vessel built downward, drilling mud as the control mechanism, and the mud weight window as the vice that makes wells telescopic.Formation fluids are pressurised by the weight of overlying rock, so an open hole to the surface invites a blowout; control comes not from a valve but from a kilometres-tall column of drilling mud, weighted with barite so its bottom-hole pressure slightly exceeds pore pressure. Mud simultaneously cools the bit, carries cuttings up to the shale shakers where a geologist reads them as the well's only news, and deposits a filter cake that holds the hole open. Too light and formation fluid enters as a kick — a runaway with gas, since a rising bubble expands and pushes mud out, lowering pressure further; too heavy and you exceed the fracture gradient and lose circulation, which drops the column height and can trigger the very kick you feared. The window narrows with depth, so wells are telescopic: drill until it closes, run and cement casing to seal that formation away, then continue with a smaller bit.
- SteeringExplain directional and horizontal drilling as a contact-length argument, and the mud motor, MWD/LWD and mud-pulse telemetry as what makes geosteering possible.A reservoir is a bed sheet — twenty metres thick and kilometres across — so a vertical hole contacts twenty metres of pay while a lateral along the layer can contact two thousand, a hundredfold difference from the same rock. The mud motor makes this ordinary by putting a turbine in the bottom-hole assembly so mud flow spins the bit while the pipe stays still, allowing a slight bend just above the bit: hold the pipe to slide and curve, rotate it to average the bend out and go straight. Measurement-while-drilling computes inclination and azimuth behind the bit while logging-while-drilling reads gamma, resistivity and density of the rock beside it, and because no wire can pass a rotating string the data is telegraphed up by pulsing a valve to modulate mud flow at a few bits per second — enabling geosteering, where a geologist keeps a bit inside a twenty-metre bed kilometres down, in real time, with a few seconds of lag.
- The Well as an InstrumentUnderstand well logging as interrogating unseen rock through physical proxies, why resistivity finds the oil, and why only cores and drill stem tests end the inference.Cuttings are crushed and late and seismic resolves only tens of metres, so the borehole itself becomes the instrument: wireline tools measure the rock through the hole's wall, and no single log says 'oil'. Gamma ray is a shale detector because clays concentrate potassium, thorium and uranium; density, neutron and sonic logs each reach porosity by different physics; and resistivity finds the hydrocarbons because rock and oil insulate while salty formation water conducts, so high resistivity in a demonstrably porous interval means the pores hold something other than brine — with Archie's equation turning that into a saturation figure. Logs remain weak on permeability, the number that decides everything, so ground truth comes from cutting a core or from a drill stem test that makes the reservoir flow, with rate indicating permeability and the pressure build-up revealing reservoir size and boundaries.
- Buying the Permeability You Weren't GivenEstablish that hydraulic fracturing is a permeability technology rather than a discovery, and explain perforation, fracture initiation, proppant and staging.Shale is the source rock where all oil is made, and a great deal never migrated out — but its permeability of 10⁻⁸ to 10⁻⁴ millidarcys, the nanodarcy range, is a million to a billion times below the millidarcy permeabilities of conventional reservoirs and roughly eight orders of magnitude short of the ~100 md producibility threshold. So fracturing found no oil: everyone knew it was there, and what changed was permeability, which is built rather than found. Shaped charges perforate the cased and cemented lateral; fluid is pumped past the fracture gradient the driller spent the whole well avoiding, splitting the rock perpendicular to the least compressive stress; proppant — sand or ceramic beads — washes in and holds the crack open against two kilometres of overburden; and dozens of isolated stages repeat this along the lateral. The matrix is unchanged and still barely flows — the oil simply no longer has far to travel.
- The Decline CurveExplain Arps decline curves and the b factor, derive shale's steep decline from the fracture-plus-matrix geometry, and recognise the treadmill that turns exploration into manufacturing.Every well declines from its first day because producing lowers reservoir pressure and therefore flow; in 1945 Jan Arps published three empirical equations — fit to data rather than derived from physics — with the general hyperbolic form q(t) = qᵢ/(1 + b·Dᵢ·t)^(1/b), where b = 0 gives exponential decline, b = 1 harmonic, and values between give hyperbolic, and integrating the fitted curve yields the estimated ultimate recovery the business is priced on. Shale broke the shape: rock beside the propped fractures drains fast, then the well waits on nanodarcy matrix seepage, so tight gas and shale wells show extended transient flow fit with b > 1 — outside Arps's intended range, and why practitioners are told to use PLE, SEPD or Duong models instead. Steep decline forces constant drilling to hold output flat, replacing geological risk with industrial risk and turning a shale play into a manufacturing business competing on cost per well.
- Offshore: Everything Becomes LogisticsUnderstand offshore development as the point where logistics rather than geology decides, and explain dynamic positioning, the riser, seabed BOPs and subsea tiebacks.The rock does not know it is underwater — traps, mud windows and logs are unchanged — but every tonne and every person arrives by ship or helicopter, so the cheapest possible offshore action exceeds a good onshore field's whole development and a field must be enormous before anyone starts. Fixed steel jackets work only to a few hundred metres, beyond which vessels must float and hold station by dynamic positioning: GPS plus seabed acoustic beacons driving thrusters to keep a 100,000-tonne ship within metres of a point for months. The riser, hanging through kilometres of moving water, is the well's most exposed organ, which is why the blowout preventer sits on the seabed so the well can be shut in if the vessel disconnects — and the most elegant answer is the subsea completion, a seabed Christmas tree tied back by pipeline to a distant platform, so nobody is ever above the well.
- The Honest LedgerPresent the industry's honest ledger — energy density, methane leakage, induced seismicity, water, well integrity — attributing contested measures and declining to adjudicate.Hydrocarbons' density, stability, pumpability and on-demand availability are why transport, aviation, shipping and industrial heat are hard to decarbonise and why these molecules are feedstock as well as fuel — an argument about difficulty, not a defence. On methane, Alvarez et al. (Science, 2018) estimated U.S. supply-chain emissions at 13 ± 2 Tg/yr for 2015, about 2.3% of gross gas production and roughly 60% above the EPA inventory, attributing the gap to inventory methods missing abnormal operating conditions, with genuine methodological dispute between top-down and bottom-up measurement. On seismicity, Oklahoma's rate increase is largely attributed by USGS work to deep wastewater disposal rather than fracturing itself — the 2016 M5.8 Pawnee earthquake being the largest known induced by disposal — because sustained large-volume injection raises pore pressure until an already-loaded fault slips, the same mechanism as geothermal EGS. Each item is measured and argued, and the disputes that look factual are usually about weighing.
Questions this course answers
Pumice is full of holes — high porosity. Why would it make a catastrophic oil reservoir?
Porosity says how much oil is in there; permeability says whether it can ever get out, and nothing forces them to go together. A rock riddled with isolated pores is like a chocolate bar full of bubbles — every drop locked in a private cell forever. Keeping these two apart is most of the intuition in this subject.
A rock needs permeability above roughly 100 md to be an exploitable reservoir without stimulation — and rocks significantly below 100 md make efficient seals. What follows from those two facts together?
The same property, on the same scale, decides two opposite things — which is the joke the Earth is playing. Every field is a collaboration: no seal and the oil left long ago; no permeability and the oil is there but you cannot have it. So what you're hunting isn't a substance, it's an arrangement — and it's invisible, kilometres down.
Oil is buoyant and spends millions of years migrating upward from the shale that made it. What does that imply about the tar pits and oil seeps humans have known for millennia?
Oil is always trying to leave, and almost all of it has succeeded — which is why a trap is the rarest item on the shopping list. A seal alone isn't enough either: without a closing geometry, the seal merely diverts the migrating oil sideways rather than stopping it. Every field that exists is oil that got stuck.
A seismic survey finds a boundary between two very different rock types completely invisible — no echo at all. How is that possible?
R = (Z₂ − Z₁)/(Z₂ + Z₁), so when the impedances match, R is zero and there is no echo regardless of how different the rocks are geologically. This is why a seismic survey doesn't image rock — it images contrasts. A hundred metres of uniform sandstone is a blank; all the signal lives at the interfaces.
The equation t = 2d/V seems to convert echo time straight into depth. Why doesn't that solve seismic imaging?
The circularity is the discipline. To migrate an echo back to its true position you need the velocity model; to build the velocity model you need a decent image. It's resolved iteratively — guess, migrate, judge whether the image makes geological sense, adjust, repeat — which is why oil companies have been among the largest civilian owners of supercomputers. They're focusing, not simulating.
Why do interpreters treat bright spots with caution despite the sound physics behind them?
The very sensitivity that makes a bright spot a direct hydrocarbon indicator is what makes it unable to tell you how much. When these were first recognised they looked like the end of exploration risk; then people drilled them. It's the honest summary of seismic generally — an inference built by processing that assumed things, with the well as the only source of truth.
Grounded in trusted sources
- Wikipedia — Permeability (Earth sciences) (~100 md threshold to produce without stimulation; below that rocks form efficient seals; 1 darcy ≈ 10⁻¹² m²; oil reservoir rocks 10–1,000 md; fresh granite 0.001–0.01 md): https://en.wikipedia.org/wiki/Permeability_(Earth_sciences)
- Wikipedia — Reflection seismology (Z = vρ; R = (Z₂ − Z₁)/(Z₂ + Z₁); vibroseis and dynamite; marine air guns at 2,000 psi, 2,000–7,000 in³; geophones and hydrophones; t = 2d/V; CMP stacking 'significantly lowers the random noise'; migration; bright spots and fizz gas; 3D seismic from 1960s experiments, large datasets late 1970s, widespread 1980s–1990s): https://en.wikipedia.org/wiki/Reflection_seismology
- Wikipedia — Petroleum geology (source rock, migration, reservoir, seal and trap): https://en.wikipedia.org/wiki/Petroleum_geology
- Wikipedia — Petroleum trap (anticlinal, fault, salt and stratigraphic traps): https://en.wikipedia.org/wiki/Petroleum_trap
- Wikipedia — Drilling fluid (cooling, cuttings transport, filter cake, hydrostatic control; barite): https://en.wikipedia.org/wiki/Drilling_fluid
- Wikipedia — Blowout (well drilling) (kicks; gas expansion; blowout preventers): https://en.wikipedia.org/wiki/Blowout_(well_drilling)
- Wikipedia — Directional drilling (mud motors; sliding and rotating; rotary steerable systems): https://en.wikipedia.org/wiki/Directional_drilling
- Wikipedia — Measurement while drilling (MWD/LWD; mud pulse telemetry): https://en.wikipedia.org/wiki/Measurement_while_drilling
Every Wunder lesson is built from real, reputable sources — never invented.
Related Science courses
Wunder is a personalized learn-anything platform — tell it any topic and it builds a beautiful, fact-checked course in minutes, with narration, a knowledge check, and a college-style University track.
Browse more Science courses · All topics · Home
© 2026 Wunder Learning LLC · Terms & Privacy