Multilateral wells

Quick refresher

A multilateral well is a main bore and several laterals with a common wellhead. The laterals can be in different formations, one above another. Or they can be in one formation, in different directions or as a “fishbone” (laterals branching off a horizontal wellbore like the bones of a fish).
The junction must support the rock, isolate the bores if necessary, and give tools access to each. The TAML classification (the industry's classification of junctions by complexity) distinguishes six levels: from an open bore with no support to a junction that is pressure-tight by virtue of the casing itself.
Laterals in one formation share the drainage area. In the example in this lesson, two 500 m bores give not twice but 1.2–1.6 times as much as one, depending on the angle between them. The gain is greater when a larger share of the resistance falls on the zone right at the wellbore, and when the area allotted to the well is smaller.
Laterals in different formations do not interfere in the reservoir, but they work on a common bottomhole. With different reservoir pressures, crossflow goes between them through the wellbore, as in commingled production.
In a low-permeability formation the laterals work independently for a long time, and a multilateral well first of all speeds up withdrawal. It adds reserves only if the laterals reach what a single wellbore would not drain.
The benefit is the common upper wellbore, casing strings and wellhead. The costs are the junction, the time to build it, and risk. A multilateral well pays off when the upper part of the well is expensive: offshore, on cramped well pads, or at great depth.

Key relationships

J = \dfrac{2\pi k_h h}{\mu B\,R}

productivity in terms of the resistance to inflow $R$ — for a single wellbore and for a system of laterals

G_N = \dfrac{J_N}{J_1} = \dfrac{R_1}{R_N}

gain from $N$ laterals in one formation against one; without interference it would be $N$

R = \dfrac{\Delta H}{\sin\theta_2 - \sin\theta_1},\qquad i = \dfrac{1718.9}{R}

radius (m) and build-up rate (°/30 m) from a junction with inclination $\theta_1$ to $\theta_2$ over a vertical depth $\Delta H$

t \approx \dfrac{(d/2)^2}{4\eta}

when parallel laterals a distance $d$ apart begin to interfere (order of magnitude)

Common pitfalls

Expecting a second lateral to double the rate. In one formation the bores share the drainage area.
Choosing a junction by price without checking whether the rock at the junction will withstand the drawdown over the whole working life.
Not thinking in advance about how to enter each lateral. Full-bore access cannot be added after the junction has been built.
Commingling laterals in formations of different pressure in a common wellbore with no way of shutting one of them off.
Comparing a multilateral well with a single one by initial rate, forgetting the common drainage area, the junction risk and the cost of workovers.

After this lesson you can

describe the main geometries of multilateral wells and the tasks for which each is suited
explain what a junction must provide, and distinguish the levels of junction complexity in the TAML classification
calculate the productivity of two laterals in one formation and in different formations, and explain why two laterals give less than two single wellbores
estimate the build-up rate (how fast the hole angle increases) from the junction to the formation, and the time after which the laterals begin to interfere
list the risks of multilateral wells, and compare the cost per unit of productivity of a multilateral well and a single horizontal well

The main bore, the laterals and the junction

The main bore, also called the motherbore, runs from the wellhead. It can be vertical, deviated or horizontal. The crew kicks off the laterals from it (starts them in a new direction), either through a window in the casing (a hole cut in the casing wall) or directly in the open hole. Each lateral is completed in its own way: left open, or fitted with a screen or a liner (bottomhole designs). The production of all the laterals is collected in the main bore and rises to the common wellhead, either through one tubing string or separately.

Russian-language literature has two names. A multi-bore well usually means a well whose laterals are kicked off above the top of the productive formation and can go to different formations. A multi-branch well (MZS, from the Russian mnogozaboynaya) is one that branches in the lower part of the main bore, already within the formation. The usage is not uniform, and both words are often used as a translation of the English “multilateral”. In this lesson “multilateral well” is used in the broad sense.

A multilateral well is different from sidetracking. In sidetracking, we drill a new wellbore from an old well, and we abandon the old wellbore below the window or leave it idle (whipstock and sidetracking). In a multilateral well all the bores work at once. In essence it is the next step after the horizontal well (Horizontal wells). The aim is the same, more contact with the formation, but there is no new upper wellbore for each horizontal section. Most often there are two or three laterals.

Where a multilateral well gives a gain

Conditions: Several formations separated by impermeable beds
Geometry: laterals one above another
Source of the gain: each formation is drained by its own horizontal lateral

Conditions: Thin or low-permeability formation
Geometry: in different directions, “fishbone”
Source of the gain: more contact with the formation per well

Conditions: Viscous oil
Geometry: several long laterals
Source of the gain: the rate grows with the length of contact with the formation; the wellbore diameter is under a logarithm and has almost no effect

Conditions: Natural fractures of unknown direction
Geometry: laterals at different azimuths
Source of the gain: at least one will cross the fractures at a favourable angle

Conditions: Fault blocks, lenses, small accumulations
Geometry: a lateral into each block
Source of the gain: one wellbore from the wellhead to several small volumes

Conditions: Mature field
Geometry: a new lateral from an old well
Source of the gain: access to undrained areas without a new wellbore from the wellhead

Conditions: Offshore, cramped well pads
Geometry: any
Source of the gain: fewer slots, pads and wellhead equipment

Conditions	Geometry	Source of the gain
Several formations separated by impermeable beds	laterals one above another	each formation is drained by its own horizontal lateral
Thin or low-permeability formation	in different directions, “fishbone”	more contact with the formation per well
Viscous oil	several long laterals	the rate grows with the length of contact with the formation; the wellbore diameter is under a logarithm and has almost no effect
Natural fractures of unknown direction	laterals at different azimuths	at least one will cross the fractures at a favourable angle
Fault blocks, lenses, small accumulations	a lateral into each block	one wellbore from the wellhead to several small volumes
Mature field	a new lateral from an old well	access to undrained areas without a new wellbore from the wellhead
Offshore, cramped well pads	any	fewer slots, pads and wellhead equipment

The summary is qualitative; the gain of each scheme is checked by calculation and by the experience of neighbouring wells.

Geometries of the laterals

SchematicThe main geometries of multilateral wells (schematics, not to scale). One cross-section shows laterals one above another in two formations. Three plan views (seen from above) show two laterals in opposite directions, a fan of several laterals in one formation, and a “fishbone”, which has short ribs off a horizontal wellbore. The orange dots are the junctions.

Stacked: The main bore is vertical or deviated, and the laterals are in different formations or at different heights of a thick formation. This suits formations separated by impermeable beds. In a thick tight formation, laterals at different heights do not feel one another for a long time (the example of interference in time is below).
Dual-opposed: Two laterals from one junction go in opposite directions. In effect this is a horizontal wellbore of double length, drilled from the middle. The reach of each lateral from the junction is half as great, and the flow in each lateral passes along only its own half of the length. So friction losses are smaller (friction in the wellbore).
Fan (planar): Several laterals in one formation, at an angle to one another. In a fractured formation, a fan protects against a mistake in the direction of the fractures, because at least one lateral will cross them at a large angle. In a tight formation with no fractures, a fan simply adds contact with the formation, but the gain depends on the angle between the laterals (see the section on productivity).
“Fishbone” (herringbone): Short laterals, called ribs, branch off on both sides of a horizontal main bore. This increases contact with a thin or tight formation. The ribs are usually left open, and tools cannot be run into them.
A lateral from an old well: In a mature field, a short lateral is drilled into a layer that the old wellbore hardly drained and where oil remains. The crew drills it through a window in the casing or directly through the tubing, with a coiled tubing unit (a unit that unreels a continuous pipe) instead of a rig, without pulling the tubing string. Such a lateral is cheaper but narrow.

Laterals in one formation share its volume, so the gain depends on their positions relative to each other (see the section on productivity below). Laterals in different formations do not share volume, but they do share the wellbore.

The junction and the TAML levels

A junction does up to three jobs. The first is mechanical support: the rock at the branching point must not collapse and block the passage. The second is hydraulic isolation: the bores and the formation at the junction must not be connected except through the intended channels, and the junction must hold a pressure difference. The third is access: instruments, coiled tubing and tools must be able to enter the required bore, for surveys, treatments and water shut-off.

In 1997, operators and service companies that had joined together in the TAML group (Technology Advancement of Multilaterals) agreed to classify multilateral wells by the design of the junction (TAML levels). The higher the level, the more reliable the junction is, mechanically and hydraulically. But its cost, its complexity and its risk in construction are higher too.

Levels of junction complexity under the TAML classification

Level: 1
Main bore: open
Lateral at the junction: open
What the junction gives: support is the rock; neither isolation nor a reliable entry

Level: 2
Main bore: cased
Lateral at the junction: open
What the junction gives: a window and an entry into the lateral; no isolation

Level: 3
Main bore: cased
Lateral at the junction: liner without cement
What the junction gives: the liner is tied to the casing, access is provided; no isolation

Level: 4
Main bore: cased
Lateral at the junction: liner cemented in place
What the junction gives: support is the pipe and cement; holds a pressure difference weakly

Level: 5
Main bore: cased
Lateral at the junction: as at levels 2–4
What the junction gives: pressure integrity from packers and tubing

Level: 6
Main bore: cased
Lateral at the junction: cased
What the junction gives: pressure integrity from the casing itself

Level	Main bore	Lateral at the junction	What the junction gives
1	open	open	support is the rock; neither isolation nor a reliable entry
2	cased	open	a window and an entry into the lateral; no isolation
3	cased	liner without cement	the liner is tied to the casing, access is provided; no isolation
4	cased	liner cemented in place	support is the pipe and cement; holds a pressure difference weakly
5	cased	as at levels 2–4	pressure integrity from packers and tubing
6	cased	cased	pressure integrity from the casing itself

“Cased” means that a casing string has been run and cemented. We build a level 5 junction on top of a level 2–4 junction by adding completion equipment: a packer above the junction and a packer in the lateral. Level 6 is rarely used. One of its varieties is a factory-made casing splitter.

SchematicJunctions of TAML levels 1–6 (schematics, not to scale). The main bore is vertical on the left, and the lateral goes down and to the right. From level to level the following are added: the casing of the main bore with a window, the liner of the lateral, cement around it, packers with tubing, and finally an integral metal casing junction.

The main consequence of the classification is the drawdown that the junction will withstand. At levels 1–3 the rock bears the pressure difference at the junction, so such junctions are suitable only for strong rock. At level 4 the cement sheath bears it. The junction then holds in weak rock as well, but, like the junctions of levels 1–3, only at moderate drawdowns. Junctions of levels 5 and 6 are separated from the formation by metal and seals, and they withstand large pressure differences. Only they stop the flows from the main bore and the lateral from mixing at the junction itself. With a junction of levels 2–4, we can separate the flows above the junction with a dual-tubing arrangement (two tubing strings). Otherwise the production mixes in the main bore (commingled production).

We choose the level by the strength of the rock, the drawdown, the need for isolation and access, and by cost. For heavy oil in strong sandstones, and for “fishbone” wells, a junction of levels 1–3 is often enough. In weak sandstones, where we need sand control (Sand control methods), the junction must be cased and cemented, which means level 4 or higher. At large drawdowns it must also be pressure-tight. We set the requirements for access before drilling, because we cannot add full-bore entry into a lateral after the junction has been built.

How a lateral is started. In an open hole, the crew kicks off a lateral from a cement plug or directly from the wall. In a cased hole, they kick it off through a window milled along a whipstock (whipstock), which is a wedge that steers the milling tool sideways. The window can be prepared in advance: a pipe with a ready-made window, closed by an easily drillable sleeve, is placed in the casing. To re-enter the required lateral later, the crew sets deflecting devices that direct the tool into the window, or uses orientable coiled-tubing assemblies.

Worked exampleHow high above the formation to place the junction

Given (assumed data)

Main bore: inclination at the junction 50°
Lateral: must become horizontal (90°) at the top of the formation
Options: the junction 60 m or 20 m (vertically) above the point of entry into the formation

Solution

Radius of the build arc from 50° to 90°: $R = \Delta H/(\sin 90^\circ - \sin 50^\circ) = 60/(1 - 0.7660) = 256.4$ m; build-up rate $1718.9/256.4 = 6.7$ °/30 m.
Junction at 20 m: $R = 20/(1 - 0.7660) = 85.5$ m, build-up rate $1718.9/85.5 = 20.1$ °/30 m.
Horizontal departure along the build arc $R\,(\cos 50^\circ - \cos 90^\circ) = R\cdot 0.6428$ : 164.8 and 55.0 m; length of the arc along the hole $R\cdot 40^\circ\cdot\pi/180$ : 179.0 and 59.7 m.

Answer: A junction 60 m above the formation requires a build-up rate of 6.7 °/30 m. A junction 20 m above it requires 20.1 °/30 m.

What it means. The closer the junction is to the formation, the steeper we have to build the angle. The second option is already at the upper limit of the medium radius (profiles by build radius, Directional wells: applications and profile). A steep build-up limits the length and the stiffness of what will pass into the lateral: a liner, a screen, instruments. A high junction, on the other hand, lengthens the section of the lateral in the seal (the cap rock) and the overlying rocks. That means more drilling, and a risk of instability where the lateral is not yet in the formation. Three groups agree on the height of the junction: geology (where the formation and the seal are), drilling (what build-up rate the tools and the liner can take) and development (where the wellbore is needed). The data in this example are assumed for illustration.

Productivity of several wellbores

Each lateral is a horizontal wellbore, and the formulas of the lesson on horizontal wells give its productivity on its own (inflow to a horizontal well). The question is how the laterals add up. The answer is different for laterals in different formations and for laterals in one formation.

Laterals in one formation drain a common volume. Each lowers the pressure not only at itself but also around its neighbour, so their drawdown cones (the funnels of lowered pressure around them) overlap (superposition in space). The resistance to inflow to a horizontal wellbore has two parts. The near part is the flow converging on the wellbore in the vertical direction, together with the skin. The far part is the flow through the formation to the wellbore in the horizontal plane. A second lateral almost halves the near part, because the flow is divided between two wellbores. But it changes the far part little, because fluid from the edge of the area still has to travel the same path. So the gain is less than two. It is greater when the near part is a larger share of the resistance.

Worked exampleTwo laterals in opposite directions against one

Given (assumed data)

Formation and oil: from the lesson on steady-state flow: $k_h$ = 60 mD, $h$ = 18 m, $\mu$ = 1.8 mPa·s, $B$ = 1.25, $r_w$ = 0.108 m; $k_v/k_h$ = 0.2, as in the lesson on horizontal wells
Drainage area: radius $r_{eh}$ = 1000 m
Options: one lateral of 500 m; two laterals of 500 m from one junction in opposite directions — one straight wellbore of 1000 m; a vertical well for comparison

Solution

Anisotropy index: $I = \sqrt{1/0.2} = 2.236$ , and $I h = 40.25$ m. The vertical term for 500 m is $\dfrac{40.25}{500}\ln\dfrac{40.25}{0.108\cdot 3.236} = 0.0805\cdot 4.7464 = 0.3821$ . For 1000 m it is $\dfrac{40.25}{1000}\cdot 4.7464 = 0.1910$ .
Horizontal terms: for 500 m the semi-axis is $a = 250\sqrt{0.5 + \sqrt{0.25 + (2000/500)^4}} = 1015.7$ m, and the term is $\ln\dfrac{1015.7 + \sqrt{1015.7^2 - 250^2}}{250} = 2.0795$ . For 1000 m, $a = 1064.3$ m and the term is $\ln\dfrac{1064.3 + \sqrt{1064.3^2 - 500^2}}{500} = 1.3882$ .
Resistances: $R_1 = 2.0795 + 0.3821 = 2.4616$ and $R_2 = 1.3882 + 0.1910 = 1.5792$ . For the vertical well, $\ln(1000/0.108) = 9.1334$ .
Productivity: $J = 0.536\cdot 60\cdot 18/(1.8\cdot 1.25\cdot R) = 578.88/(2.25\,R)$ . One lateral gives 104.5, two give 162.9, and the vertical well gives 28.2 m³/(d·MPa).
Gain from the second lateral: $162.9/104.5 = 1.56$ instead of 2. Two separate wellbores, each in its own area of the same size, would give $2\cdot 104.5 = 209.0$ .

Answer: Two laterals of 500 m give 162.9 m³/(d·MPa). This is 1.56 times one lateral and 5.8 times a vertical well.

What it means. The second lateral halved the near, vertical resistance (from 0.3821 to 0.1910). But the far, horizontal resistance fell only from 2.0795 to 1.3882, because both wellbores drain the same area of radius 1000 m. If the second lateral drained its own, separate area, its productivity would add fully to that of the first. But that is a question of hydrocarbons in place, not of rate. The data in this example are assumed for illustration.

We cannot calculate laterals at another angle with a direct formula, because their drawdown cones overlap in a more complicated way. A calculation with the superposition model (see the advanced section) gives the gain of two laterals as a function of the angle between them.

CalculatedCalculated gain of two laterals of 500 m against one in the formation of the example, as a function of the angle between them, for a drainage radius of 1000 m and of 2000 m. The dot shows laterals in opposite directions (180°). The dashed line is the gain without interference.

Angle. The smaller the angle, the more the cones overlap. At 30° the second lateral adds only about a quarter of the productivity of one lateral. At 90° it adds about 45%, and in opposite directions about 55%.
Area. The larger the area allotted to the well, the smaller the gain. In a large area the far resistance weighs more, and the second lateral hardly reduces it.
Anisotropy and skin. The smaller $k_v/k_h$ and the greater the skin, the larger the share of the near resistance, and the greater the gain. For two opposed laterals in the formation of the example, the gain grows from 1.53 at $k_v = k_h$ to 1.66 at $k_v/k_h$ = 0.02.

Worked exampleHow much the third and further laterals add

Given (assumed data)

Formation and area: from the previous example; one lateral of 500 m — $J_1$ = 104.5 m³/(d·MPa)
Options: 2, 3, 4 and 6 laterals of 500 m from one junction at equal angles (180°, 120°, 90° and 60°)
Calculation: superposition model (the advanced section)

Solution

Gain against one lateral: two give 1.55, three give 1.87, four give 2.09, and six give 2.36. For two laterals the model gives 1.554, and the standard formula in the previous example gives $162.9/104.5 = 1.559$ . The difference is 0.3%.
The increments in fractions of one lateral: the second adds $1.55 - 1 = 0.55$ , the third adds $1.87 - 1.55 = 0.32$ , the fourth adds $2.09 - 1.87 = 0.22$ , and the fifth and sixth together add $2.36 - 2.09 = 0.27$ .
Productivity of six laterals: $2.36\cdot 104.5 = 247$ m³/(d·MPa).

Answer: Six laterals are 2.36 times as productive as one, not 6 times. Each next lateral adds less than the previous one.

What it means. This is diminishing returns: the wellbores crowd into one area, and each new one takes inflow from its neighbours. If laterals are expensive, the best number is small, often two or three. The picture is different when the laterals go into different formations or blocks. It is also different in a tight formation in the first months and years, before the drawdown cones have merged (see the example below). The data in this example are assumed for illustration.

Laterals in different formations. If an impermeable bed separates the formations, the laterals do not share volume, and their productivities add fully. But they share the wellbore. The bottomhole pressure is common to them, as for formations in commingled production (common bottomhole and crossflow).

Worked exampleLaterals in two separated formations

Given (assumed data)

Lower formation: the formation of the previous examples: a lateral of 500 m — $J_1$ = 104.5 m³/(d·MPa); reservoir pressure 18.0 MPa
Upper formation: $k_h$ = 30 mD, $h$ = 9 m, $k_v/k_h$ = 0.3, the same oil, area $r_{eh}$ = 1000 m; partly depleted — 15.0 MPa
Well: a lateral of 500 m in each formation; the production mixes in the main bore; pressures are referred to the depth of the junction

Solution

Upper formation: $I = \sqrt{1/0.3} = 1.826$ , and $I h = 16.43$ m. The vertical term is $\dfrac{16.43}{500}\ln\dfrac{16.43}{0.108\cdot 2.826} = 0.1310$ . The horizontal term is 2.0795, as in the first example. The resistance is $R = 2.2105$ .
Productivity of the lateral in the upper formation: $J_{\text{up}} = 0.536\cdot 30\cdot 9/(2.25\cdot 2.2105) = 29.1$ m³/(d·MPa). For the whole well it is $104.5 + 29.1 = 133.6$ , because the formations do not share volume, and the productivities add.
A vertical well through both formations: $J = 0.536\cdot (30\cdot 9 + 60\cdot 18)/(2.25\cdot 9.1334) = 35.2$ m³/(d·MPa).
Shut-in: the pressure in the wellbore is $p_{ws} = (104.5\cdot 18.0 + 29.1\cdot 15.0)/133.6 = 17.35$ MPa. The crossflow from the lower formation into the upper one is $29.1\cdot (17.35 - 15.0) = 68.4$ m³/d.
The upper formation starts to work only when the bottomhole pressure drops below 15.0 MPa. At that point the lower one is already giving $104.5\cdot 3.0 = 313.5$ m³/d.

Answer: The well with two laterals is 3.8 times as productive as a vertical well. But with a pressure difference of 3 MPa, in the shut-in well the lower formation flows into the upper one at about 68.4 m³/d. At a withdrawal of less than 313.5 m³/d, the upper formation does not work.

What it means. Separated formations do not interfere with each other in the reservoir, and their productivities add fully. But laterals in formations with different pressures are the same as commingled production, with all its consequences. The depleted formation stands idle at a low withdrawal, and it takes in foreign fluid when the well is shut in. We can avoid this with a level 5 or 6 junction with separate channels, or with valves at the entry of each lateral. The data in this example are assumed for illustration.

Interference in time. Everything said about the gain applies to steady-state inflow. Right after start-up, each lateral works like a single wellbore, because its drawdown cone has not yet reached its neighbour. The time for the cones to merge grows with the square of the distance between the wellbores, and it is inversely proportional to the hydraulic diffusivity (how fast a pressure change spreads through the rock) (Line-source solution and the pressure function, Flow regimes).

Worked exampleWhen the laterals start to interfere

Given (assumed data)

Laterals: two parallel laterals 300 m apart
Formation: $\phi$ = 0.15, $\mu$ = 1.0 mPa·s, $c_t = 1.5\cdot 10^{-3}$ MPa⁻¹; permeability 50 mD or 0.05 mD

Solution

Hydraulic diffusivity: $\eta = 3.553\cdot 10^{-3}\,k/(\phi\mu c_t)$ (in m²/h, with $k$ in mD). It is 789.6 m²/h at 50 mD and 0.790 m²/h at 0.05 mD.
The cones merge when the radius of investigation $\sqrt{4\eta t}$ (how far the pressure change has spread) reaches the midpoint between the wellbores, 150 m. Then $t = 150^2/(4\eta)$ , which is 7.1 h at 50 mD, and 7120 h (about 297 days) at 0.05 mD.

Answer: In a 50 mD formation the laterals begin to interfere after a few hours. In a 0.05 mD formation they begin after about ten months.

What it means. In a high-permeability formation, interference arrives almost at once, so we have to calculate the steady-state gain. In a tight formation the laterals work like separate wells for a long time. A multilateral well then first of all speeds up withdrawal, and the economics of acceleration decide how many laterals are justified (value of acceleration). The estimate gives only the order of magnitude, because the pressure at the neighbour starts to change earlier, only weakly (Line-source solution and the pressure function). The data in this example are assumed for illustration.

Risks of multilateral wells

The junction. Rock at an open junction can collapse and block the lateral. The cement of a level 4 junction cracks under a large pressure difference. The seals of level 5–6 junctions wear. Failure of a junction can cost the well a lateral, or even all its production.
Clean-up. After drilling, the laterals are cleaned of drilling fluid unevenly. The flow goes where the resistance is lower, and a lateral that is cleaned less well may never reach full performance (Formation damage and damage skin).
Water and gas. Breakthrough of water or gas into one lateral raises the bottomhole pressure and takes inflow from the others, like a watered-out formation in commingled production (watering-out of one formation). We can shut off a lateral only if the junction and the completion allow it. The water crest under each lateral grows in the same way as under a horizontal well (water and gas cresting).
Access and monitoring. We can find out which lateral is producing water only by surveying each one, and for that we must enter it. Treatments, isolation and fishing in laterals are more difficult and more expensive than in a single wellbore.
Drilling. While the crew drills the second lateral, the first is open. This can cause losses (drilling fluid lost into the rock) into one bore and a kick (an unwanted inflow of formation fluid) from the other, and difficult well killing. Tools can also get stuck at the window and in a steep build-up section.
Concentrated risk. There is one well instead of two or three. If it fails, more production is lost at once.

In Kazakhstan, multilateral wells, deviated wells and wells with a horizontal section are drilled to individual technical projects. These projects must make sure that the whole set of geophysical surveys can be carried out (the Unified Rules for the Rational and Integrated Use of Subsoil, the version in force in 2026; for project documents see State regulation of field development).

When a multilateral well pays off

A multilateral well can bring two different benefits. The first is acceleration: the same reserves are produced faster, and the money arrives sooner. The second is an increase in recoverable reserves: a lateral opens a volume that would not otherwise be produced within the life of the well, such as a separate formation, a block behind a fault, an undrained area or a lens. If a lateral lies in a part of the formation that is already being drained, the benefit comes down to acceleration. The value of acceleration is limited. Even instantaneous withdrawal would not recover more than discounting (lowering the value of future money to today's value) takes away from the reserves now (acceleration of an exponential decline).

The savings come from the parts that the laterals share: the wellhead and the site (offshore, a platform slot or a subsea cluster, Field development concepts: surface facilities), the upper wellbore and its casing strings. The junction, the time spent on windows, whipstocks and entering the laterals, and the risk all add costs. The first check is the cost per unit of productivity, as for a horizontal well (will a horizontal well pay off). In the end we compare the net present value (NPV) of the options (Net present value (NPV)).

Worked exampleDoes the second lateral pay off

Given (assumed data)

Single horizontal well: cost 100 assumed units, of which the upper part down to the junction is $u$ and the lateral $100 - u$ ; productivity 104.5 m³/(d·MPa) (the first example of the section on productivity)
Well with two laterals in opposite directions: the same upper part, two laterals, the junction 15 units; productivity 162.9 m³/(d·MPa)
Risk: with a probability of 10% the junction fails, and the well is left with one lateral

Solution

Cost of the well with two laterals: $u + 2(100 - u) + 15 = 215 - u$ . At $u$ = 60 (onshore, moderate depth) it is 155 units, and at $u$ = 80 (offshore) it is 135 units.
Cost per unit of productivity: a single well costs $100/104.5 = 0.957$ . A well with two laterals costs $155/162.9 = 0.952$ at $u$ = 60 and $135/162.9 = 0.829$ at $u$ = 80.
Break-even by this measure is when $(215 - u)/162.9 = 100/104.5$ , that is, at $u = 215 - 100\cdot 162.9/104.5 = 59$ .
With the risk of junction failure, the expected productivity is $0.9\cdot 162.9 + 0.1\cdot 104.5 = 157.1$ , and the threshold rises to $u = 215 - 100\cdot 157.1/104.5 = 65$ .

Answer: The second lateral pays for itself in productivity when the upper part of the well costs more than 59% of the price of a single well. With the junction risk allowed for, it must cost more than 65%.

What it means. Onshore, with an inexpensive upper wellbore, a well with two laterals gains almost nothing in this formation, because the laterals share one area. Offshore, where the slot and the upper wellbore are the main part of the price, the gain is noticeable. If a lateral reaches reserves that a single wellbore cannot, the estimate becomes more favourable. If the junction requires level 5 or 6, the junction itself becomes more expensive. In the end we compare the NPV of the options, with production, workovers and risks over the whole life. The costs, shares and probability of failure are assumed for illustration.

Pays off more often: on offshore platforms and subsea clusters with a limited number of slots; on cramped well pads; in deep wells with an expensive upper wellbore; in separated formations and blocks that would not have paid for separate wells; in thin, tight and fractured formations; with heavy oil, where the contact area matters.
Pays off less often: in a high-permeability homogeneous formation, where the laterals begin to interfere quickly; in weak rock at high drawdowns, where an expensive pressure-tight junction is needed; where water and gas will often have to be isolated lateral by lateral; where the geology is poorly known and it is easy to put a lateral in the wrong place.

Terms and symbols

Multilateral well: A main bore and one or more laterals with a common wellhead, working at the same time. A well that branches in the formation itself is often called multi-branch.
Main bore (motherbore): The bore from the wellhead from which the laterals are kicked off.
Lateral: A side bore leaving the main one. It is usually horizontal or gently inclined in the formation.
Junction: The place where a lateral leaves the main bore. It provides support, isolation and access, depending on the level.
TAML levels: The classification of junctions from 1 (open) to 6 (pressure-tight by virtue of the casing).
“Fishbone”: Short open ribs on both sides of a horizontal main bore.
Gain from laterals $G_N$: How many times more productive $N$ laterals are than one. In a single formation it is always less than $N$ .
Whipstock: A steel wedge along which the mill cuts a window in the casing and goes into the new lateral.

Common mistakes and misconceptions

Wrong

“Two laterals, two rates.”

In fact

In one formation the laterals share the drainage area. In the example in this lesson, two 500 m bores in opposite directions gave 1.56 times as much as one lateral, and six laterals gave 2.36 times as much.

Wrong

“A junction is just a place where the wellbore forks.”

In fact

The junction decides whether the rock will withstand the drawdown, and whether we can shut off a lateral and enter it. These properties are built in when the junction is designed. We cannot add entry into a lateral after construction, and we can add pressure integrity only by a level 5 completion on top of a finished junction.

Wrong

“Laterals in different formations add up without consequences.”

In fact

In the formation, yes. In the wellbore, no. With different pressures, the formation with the higher pressure flows into the one with the lower pressure, and a depleted formation stands idle at a low withdrawal. We need separate channels or valves.

Wrong

“A multilateral well is always cheaper than two horizontal ones.”

In fact

The savings are only on the common upper part, while the junction and the risk add costs. In the example in this lesson, the gain is almost nil onshore with an inexpensive upper wellbore, and it is noticeable offshore.

Wrong

“In a tight formation the laterals interfere with each other anyway, so it makes no sense to drill them.”

In fact

Quite the opposite. In a tight formation the drawdown cones merge only after months and years, and all that time the laterals work almost like separate wells. This is the main argument for multilateral wells in low-permeability formations.

In depth: advanced detail

How the interference between laterals is calculated

We calculate laterals in one formation by adding their drawdown cones in the horizontal plane. We divide each lateral into segments of length $\Delta$ , with uniform inflow within each segment. In steady-state flow, the pressure drop that a segment creates at a point of the formation is proportional to the average of $\ln(r_{eh}/\rho)$ along the segment, where $\rho$ is the distance from a point of the segment to that point. We then add the resistance of the vertical convergence of the flow to the wellbore. This is the same term as in the standard model of a horizontal well, referred to the length of the segment.

Laterals in one formation: steady-state inflow (summation of segments)

\frac{2\pi k_h h\,(p_e - p_w)}{\mu B} = \sum_j q_j\,G_{ij} + q_i\,\frac{I h}{\Delta}\left[\ln\frac{I h}{r_w(I + 1)} + S\right],\qquad G_{ij} = \frac{1}{\Delta}\int_{\Delta_j}\ln\frac{r_{eh}}{\rho_{ij}(s)}\,\dd s

where:

$q_j$ rate of the $j$ -th segment of the lateral at standard conditions, m³/s
$G_{ij}$ contribution of segment $j$ to the pressure drop at the middle of segment $i$ , —
$\rho_{ij}(s)$ distance from the middle of segment $i$ to the point $s$ of segment $j$ , m
$\Delta$ length of a segment, m
$r_{eh}$ radius of the drainage area of the well, m
$p_e,\ p_w$ pressure at the boundary of the area and in the wellbores, Pa
$I,\ h,\ r_w,\ S$ anisotropy index, thickness, wellbore radius, skin — as in the standard model, —, m, m, —
$k_h,\ \mu,\ B$ horizontal permeability, viscosity and formation volume factor of the oil, m², Pa·s, m³/m³

Units: SI; $\ln$ is the natural logarithm

Holds when

We write the equation for the middle of each segment. All segments are at the same pressure $p_w$ , so there is no friction in the wellbores.
We take the constant-pressure boundary $p_e$ at a distance $r_{eh}$ into account only roughly, as in the standard model. For laterals diverging from the centre of the area at equal angles, the exact solution for a circle (method of images) gives the same gains to within 0.5%, although the laterals reach half the radius.
The vertical convergence to the wellbore does not overlap between different laterals. The laterals are farther from one another than $I h$ , except near the junction, where the model slightly overestimates the gain. That is why the gain curve starts at 15°.

This is a system of linear equations, one for each segment. Its solution gives the rates of the segments, and their sum divided by $p_e - p_w$ is the productivity of the system of laterals. The integral in $G_{ij}$ can be written with elementary functions.

Checks of the model. For one straight wellbore without the vertical term, the calculation tends to $\ln(4r_{eh}/L)$ , the limit of the horizontal term of the standard model for $r_{eh} \gg L$ . With the vertical term, the model gives a resistance 3% greater than the standard one (2.53 against 2.46 for a 500 m lateral in the formation of the examples). The reason is that the near resistance evens out the inflow along the wellbore, whereas the standard model adds the horizontal term of an infinitely conductive wellbore to the vertical term for uniform inflow. That is why we take only ratios (the gain) from the model, and we calculate the productivity of a single wellbore by the standard formula. For two laterals in opposite directions, the model and the standard formula agree to within fractions of a per cent. Forty segments per lateral are enough for the gain not to change in the third decimal place.

Connections to other topics

Horizontal wells — productivity of a single wellbore, skin, friction in the wellbore and the water crest.
Deflection tools and downhole motors — the whipstock and the casing window for a lateral.
Directional wells: applications and profile — radius of curvature and build-up rate.
Multiple-zone completions: commingled and segregated production — common bottomhole, crossflow and separate operation of formations.
Superposition principle — adding the drawdown cones of several wellbores.
Well completion: opening the reservoir and the tubing string — the bottomhole designs from which the completion of the laterals is chosen.
Acceleration projects and incremental analysis — what it costs to speed up the withdrawal of the same reserves.
Field development concepts: surface facilities — platform slots, subsea clusters and well pads.
Drilling time and cost — what the cost of a well and of its parts is made up of.

Check yourself

1
Two 500 m laterals have been drilled from one junction at an angle of 30° to each other in a homogeneous formation. What can be expected from the second lateral?
1. A.The same as from a lateral in the opposite direction
2. B.A reduction in productivity
3. C.A doubling of productivity
4. D.A small addition: the laterals are close together and interfere strongly
Show the answer
Correct answer: D. A small addition: the laterals are close together and interfere strongly
At a small angle the drawdown cones of the wellbores almost coincide. In this lesson's calculation, the gain is about 1.24 at 30°, about 1.44 at 90° and about 1.55 in opposite directions (drainage radius 1000 m).
2
Laterals in two formations with different pressures are to work through separate tubing strings, and one of them will have to be shut off when water breaks through. A junction of which TAML level is needed?
Show the answer
We need hydraulic isolation of the junction. That means level 5, where packers above the junction and in the lateral, with separate channels, provide it, or level 6, where the casing itself provides it. Only such junctions hold the pressure difference between the formations at the junction itself. A junction of levels 2–4 with a dual-tubing arrangement above it can also keep the flows separate. But it does not itself hold the pressure difference (at levels 2–3 the rock bears it, and at level 4 the cement bears it). When the formation pressures differ, this is a weak point.
3
Why does the second lateral give a greater gain in an anisotropic formation with low vertical permeability?
1. A.The gain does not depend on anisotropy
2. B.In an anisotropic formation the laterals do not interfere with each other
3. C.Most of the resistance is convergence to the wellbore in the vertical direction, and a second lateral almost halves it
4. D.A small $k_v$ reduces the horizontal term
Show the answer
Correct answer: C. Most of the resistance is convergence to the wellbore in the vertical direction, and a second lateral almost halves it
The second lateral divides the flow in half, so it halves the near, vertical resistance. It changes the far resistance little. The greater the share of the near resistance, the greater the gain. For two opposed laterals in the formation of the example, the gain grows from 1.53 at $k_v = k_h$ to 1.66 at $k_v/k_h$ = 0.02.
4
The junction is 40 m (vertically) above the top of the formation, and the inclination of the main bore at the junction is 60°. What build-up rate is needed for the lateral to become horizontal at the top of the formation? And if the inclination at the junction is 30°?
Show the answer
$R = 40/(1 - \sin 60^\circ) = 40/(1 - 0.8660) = 298.5$ m, $i = 1718.9/298.5 \approx 5.8$ °/30 m. This is a long radius. At 30° at the junction, $R = 40/(1 - 0.5) = 80$ m and $i \approx 21.5$ °/30 m. This is steeper than a medium radius. So it is better to put the junction where the main bore is already strongly inclined, or higher above the formation.
5
Onshore, the upper part of the well costs 40% of the price of a single horizontal well, the junction 15%, and the second lateral in this formation gives a gain of 1.5. What does the cost per unit of productivity show?
1. A.They cannot be compared
2. B.The single well is better value: 175 units for one and a half times the productivity costs more than 100 for one
3. C.The well with two laterals is better value
4. D.The options are equivalent
Show the answer
Correct answer: B. The single well is better value: 175 units for one and a half times the productivity costs more than 100 for one
The well with two laterals costs $40 + 2\cdot 60 + 15 = 175$ units and gives 1.5 times the productivity of the single well. That is $175/1.5 \approx 117$ against 100 for the same productivity. What could make it pay is an increase in reserves that a single wellbore cannot reach, or a more expensive upper part, as offshore.
6
Two parallel laterals 200 m apart in a formation with a hydraulic diffusivity of 2 m²/h. When will they start to interfere noticeably?
Show the answer
As an order of magnitude, this happens when the radius of investigation $\sqrt{4\eta t}$ reaches the midpoint between the wellbores, 100 m. Then $t \approx 100^2/(4\cdot 2) = 1250$ h, which is about 52 days. Before that the laterals work almost like separate wells. After it, their total rate falls faster than that of two independent wellbores.

Updated: 18 October 2026