Completion string equipment

Quick refresher

Each function of the tubing string has its own component. The packer isolates the annulus. The safety valve shuts off the flow. The nipple receives a plug. A sliding sleeve or a side-pocket mandrel connects the tubing with the annulus. We limit the set to what is necessary, because every component is also a place where something can fail.
A packer is held in place by slips (toothed wedges), which cones push into the wall of the casing. A rubber element seals it. A retrievable packer is released by pulling or turning the tubing. A permanent packer has to be milled out (ground away with a milling tool). A mechanical packer holds a pressure difference only in the direction in which it was set, unless it has slips that grip the other way or a hold-down (an anchor).
When pressure and temperature change, the tubing changes length. Cooling shortens it. So do pressure inside the tubing, which makes the tubing swell (ballooning), and the pressure force on the seals. Heating makes the tubing longer. Free seals slide, so they need spare stroke (room to move). Latched seals pass the force on to the packer and the tubing.
The pressure force on the seal assembly equals the pressure changes multiplied by the ring-shaped areas between the packer bore (the opening in the packer) and the inside and outside diameters of the tubing. If the packer bore is wider than the tubing, the pressure in the annulus reduces this force.
The safety valve stays open while the control line is under pressure. A spring closes it when the pressure is bled off (let out). The column of fluid in the line works against closing, so the closing pressure of the valve limits the setting depth.
A nipple has a profile (a shaped groove) for a lock and a polished bore for seals. Selective nipples are all the same size, and the lowest one is a no-go nipple (it has a shoulder that stops the tool). Every nipple narrows the bore, and narrow points wear away (erode) the tubing. So flow couplings (thick-walled short pipes) are installed above and below the nipples.

Key relationships

\Delta F_{\text{p}} = (A_{\text{p}} - A_{\text{i}})\,\Delta p_{\text{i}} - (A_{\text{p}} - A_{\text{o}})\,\Delta p_{\text{o}}

pressure force on the seal assembly in the packer (upward positive)

F_{\text{b}} = 2\nu\,(\Delta\bar p_{\text{i}} A_{\text{i}} - \Delta\bar p_{\text{o}} A_{\text{o}})

ballooning force (the swelling of the tubing under internal pressure); $2\nu = 0.6$ for steel, where $\nu$ is Poisson's ratio (how much a material gets thinner when it is stretched)

F_{\text{t}} = E\,\alpha\,A_s\,(-\Delta\bar T)

temperature: $E\alpha \approx 2.57$ MPa per °C

\Delta L = \dfrac{F\,L}{E\,A_s}

shortening of free tubing under the sum of forces $F$

p_{\text{c}} \ge p_{\text{i}}(H) + p_{\text{open}} - \rho_{\text{c}}\,g\,H

control line pressure of the safety valve at the surface

Common pitfalls

Assuming that nothing happens to the tubing once the packer is set. During a reservoir treatment, or when cold water is injected, the tubing shortens by a metre or more.
Using a packer that is set by the weight of the tubing in an injection well. Pressure from below and the shortening of the tubing take the load off it.
Forgetting that every nipple narrows the bore: a device for the lowest nipple must pass through all the upper ones.
Setting the safety valve “a bit deeper to be safe”: the column of fluid in the control line may prevent it from closing.
Having no plan for what to do if the sliding sleeve does not close after the well has been killed.

After this lesson you can

name the components that do the jobs of the tubing string, and explain why they sit in this order
compare retrievable packers (which can be pulled out) and permanent packers (which are milled out), compare the ways of setting them (fixing them in the casing), and choose a packer to suit the direction of the loads
calculate the pressure force on the seal assembly in a packer, the change in tubing length due to pressure and temperature, and the load on the packer with free and latched seals
choose a seal assembly (anchor, locator or expansion joint) to suit how far the tubing is expected to move
explain how tubing-retrievable and wireline-retrievable safety valves are built, and calculate the control line pressure (the pressure in the thin hydraulic line that holds the valve open) and the maximum setting depth
describe landing nipples, sliding sleeves and side-pocket mandrels: what is installed in them and how they fail

The tubing string from top to bottom

The lesson Well completion: opening the reservoir and the tubing string lists the functions of the tubing string: to hold pressure and carry the flow, to isolate the annulus, to shut off the flow below the wellhead, to plug the tubing, and to connect the tubing with the annulus (functions of the tubing string). Here we look at how the components that do these jobs are built, and why each one stands where it does.

SchematicThe tubing string of a naturally flowing well (a well that flows without a pump) with a permanent packer (diagram, not to scale). It runs from the tubing hanger at the wellhead to the entry guide at the lower end. Above the packer is a seal assembly with a locator. Below the packer are nipples, a perforated joint and a gauge in the lowest nipple.

Components of the tubing string: what they do and where they sit

Component: Tubing hanger, Christmas tree
What it does: carry the weight of the tubing, seal the annulus at the wellhead, control the flow
Where it sits: wellhead

Component: Safety valve
What it does: shuts off the flow if the wellhead is damaged or removed
Where it sits: tens to hundreds of metres below the wellhead

Component: Flow couplings
What it does: take the erosion that eddies cause at narrow points (restrictions)
Where it sits: above and below the valve and the nipples

Component: Side-pocket mandrel
What it does: gas lift, chemical injection, emergency circulation
Where it sits: at depths from the gas lift design

Component: Sliding sleeve
What it does: connects the tubing with the annulus for flushing and killing
Where it sits: as low above the packer as possible

Component: Seal assembly, expansion joint, disconnect device
What it does: let the tubing change length and be pulled without the packer
Where it sits: above and in the packer

Component: Packer
What it does: isolates the annulus
Where it sits: as close to the top of the reservoir as possible

Component: Landing nipples
What it does: receive a plug, a gauge, a choke, a check valve
Where it sits: above and below the packer

Component: Perforated joint
What it does: a second flow inlet if the end of the tubing is blocked by a gauge
Where it sits: above the lowest nipple

Component: Entry guide
What it does: guides wireline tools back into the tubing
Where it sits: lower end

Component	What it does	Where it sits
Tubing hanger, Christmas tree	carry the weight of the tubing, seal the annulus at the wellhead, control the flow	wellhead
Safety valve	shuts off the flow if the wellhead is damaged or removed	tens to hundreds of metres below the wellhead
Flow couplings	take the erosion that eddies cause at narrow points (restrictions)	above and below the valve and the nipples
Side-pocket mandrel	gas lift, chemical injection, emergency circulation	at depths from the gas lift design
Sliding sleeve	connects the tubing with the annulus for flushing and killing	as low above the packer as possible
Seal assembly, expansion joint, disconnect device	let the tubing change length and be pulled without the packer	above and in the packer
Packer	isolates the annulus	as close to the top of the reservoir as possible
Landing nipples	receive a plug, a gauge, a choke, a check valve	above and below the packer
Perforated joint	a second flow inlet if the end of the tubing is blocked by a gauge	above the lowest nipple
Entry guide	guides wireline tools back into the tubing	lower end

An example for a naturally flowing well with a single reservoir. The set and the order change with the method of production, the number of reservoirs (Multiple-zone completions: commingled and segregated production) and the safety requirements.

We design the tubing itself in the same way as casing (pipe strength ratings, Casing design). For tension, we use the weight of the string and the forces discussed below. For internal pressure, we usually take the wellhead of a shut-in gas well, where the pressure outside is lowest. For collapse (crushing by pressure from outside), we take the bottom, when the tubing is empty and the annulus is full. Connections in gas wells are made gas-tight. We choose the steel according to the content of carbon dioxide and hydrogen sulfide, from carbon steel to chromium and nickel alloys. Sometimes the string is made of tubing of different diameters: wider at the top, where the gas expands and the flow speeds up, and narrower at the bottom, where the liner of the casing string restricts the bore.

At the top, the string hangs on the tubing hanger in the tubing head of the wellhead. The hanger carries the weight of the tubing and the forces that come from its lengthening and shortening. It seals the annulus. It also passes the control line of the safety valve, or the pump cable, to the outside. When the Christmas tree is removed, a back-pressure (check) valve is set in the profile of the hanger, and it shuts off the tubing. The lesson Wellheads and running casing covers the design of the wellhead and the casing heads.

Packers: how they hold and how they seal

A packer is a central pipe (the mandrel), through which the flow passes, with parts mounted on it. Cones push out the slips, which have hardened teeth, and drive them into the wall of the casing. The slips keep the packer from moving. The sealing element is made of rubber packing cups. When the packer is set, it is squeezed along its axis and expands to the wall of the casing, which closes off the annulus. A lock ring with a ratchet thread (teeth that let it move one way only) keeps the set packer from releasing. At high pressures, metal back-up rings stop the rubber from being squeezed (extruded) into the gap.

SchematicOn the left, a packer while it is being run in and after setting (diagram): the cones have pushed out the slips, and the packing elements are squeezed against the wall of the casing. On the right, three ways of connecting the tubing to a permanent packer: an anchor seal assembly with a latch (the tubing does not move), a locator seal assembly in an extended bore (the seals move upwards) and an expansion joint above the packer.

A retrievable packer is run on the tubing as part of the string. We release it by reversing the setting steps: by pulling, which shears calibrated shear screws (screws made to break at a set load), or by turning the tubing. It can be pulled out and used again. But it is more complex, and it holds smaller pressure differences and loads. A permanent packer (drillable) is usually set on a separate run, either on electric line (a cable that carries power and signals) or on pipe. After that, the seal assembly on the lower end of the tubing is stabbed (pushed) into its polished bore. A permanent packer is stronger, and it holds larger pressure differences and higher temperatures. For the same outside diameter it gives a larger bore. The tubing above it can be pulled out while the packer stays in the well. We remove it by milling, a job of several hours, and then fish out (catch and pull out) the remains of the packer.

Mechanical setting: We turn the tubing and put its weight down on the packer (compression-set), or we pull on the tubing (tension-set). This is simple, but we have to move the tubing. In a deviated hole, weight and rotation reach the packer poorly. And with simple mechanical packers, we have to keep up the load of weight or tension the whole time the packer is in service.
Hydraulic setting: We set the packer with pressure in the tubing. For this, the tubing below the packer is closed off for a short time, either with a plug in a nipple or with a ball dropped onto a seat. We do not move the tubing. This is convenient in deviated wells and when there are several packers. We can assemble the string and secure it at the wellhead before setting.
Setting on electric line: We run a permanent packer on a geophysical (electric) cable with a setting tool. In the tool, a slow-burning charge makes gas pressure, and the gas pressure moves the parts of the packer. We correlate the depth with a log (match it to a well log) to within a fraction of a metre.
Inflatable packers: A reinforced rubber sleeve that we inflate with fluid. It seals over a wide range of diameters, including open hole. We use inflatable packers mainly in testing and workovers.

A mechanical packer holds a load in the direction in which it was set. A packer set in compression by the weight of the tubing seals better when the pressure difference acts from above. It can move and lose its seal when the pressure below it is higher than the pressure above it. With a packer set in tension it is the other way round. Pressure from below presses the slips harder, and pressure from above takes the load off it. That is why we use tension packers where the pressure from below is higher and the tubing shortens: in injection wells and during reservoir treatments. We use compression packers where the pressure difference points from top to bottom, for example in wells with a low reservoir pressure. If the pressure difference can change direction, we use packers with slips in both directions, packers with a hydraulic hold-down (an anchor that pressure from below presses against the casing), or permanent packers.

Which packer to choose

Conditions: Injection well, reservoir treatments
Suitable: tension-set; with slips in both directions or a hold-down; permanent
Why: pressure from below and shortening of the tubing unload a packer set by weight

Conditions: Deviated hole, several packers
Suitable: hydraulic-set
Why: no need to turn the tubing or put weight on the packer

Conditions: High pressures and temperatures, gas
Suitable: permanent
Why: holds a larger pressure difference, and loads in both directions

Conditions: Frequent workovers, temporary string
Suitable: retrievable
Why: is released without milling

Conditions: Two or three tubing strings
Suitable: dual- or triple-bore
Why: a separate bore for each string

Conditions	Suitable	Why
Injection well, reservoir treatments	tension-set; with slips in both directions or a hold-down; permanent	pressure from below and shortening of the tubing unload a packer set by weight
Deviated hole, several packers	hydraulic-set	no need to turn the tubing or put weight on the packer
High pressures and temperatures, gas	permanent	holds a larger pressure difference, and loads in both directions
Frequent workovers, temporary string	retrievable	is released without milling
Two or three tubing strings	dual- or triple-bore	a separate bore for each string

ISO 14310 sets the requirements for packers and bridge plugs (plugs that seal the casing) and their test grades. Its interstate version is GOST ISO 14310-2014.

Why tubing changes length: forces on the packer

Tubing hung in the tubing head and sealed in the packer is a long steel rod. When the pressures inside and outside the tubing, or the temperature, change, the rod tends to get longer or shorter. If the seals in the packer can slide, the tubing moves. If the seals are latched, the movement turns into a force, and the tubing, the packer and the hanger have to carry it. Four effects act (tubing movement):

Pressure force on the seals (the piston effect): the pressures in the tubing and in the annulus act on the ring-shaped end areas of the seal assembly, as they would on a piston.
Ballooning: pressure inside swells the tubing sideways, and because of the Poisson effect (a material that gets fatter gets shorter) the tubing shortens. Pressure outside squeezes the tubing in diameter and makes it longer.
Temperature: hot production makes the tubing longer, and cold injected fluid makes it shorter.
Helical buckling: under pressure the lower part of free tubing bends into a helix (a spiral) along the walls of the casing, which also shortens it (advanced section).

SchematicThe areas on which pressure acts at a permanent packer with a seal assembly (section, not to scale; dimensions are from the example). The pressure in the tubing pushes the tubing upwards. It does this through the ring between the packer bore and the inside diameter of the tubing. The pressure in the annulus pushes the tubing downwards, through the shoulder between the packer bore and the outside diameter of the tubing. The pressure difference across the packer itself acts on the ring between the casing and the bore.

Pressure force on the seal assembly in a packer

\Delta F_{\text{p}} = \left(A_{\text{p}} - A_{\text{i}}\right)\Delta p_{\text{i}} - \left(A_{\text{p}} - A_{\text{o}}\right)\Delta p_{\text{o}}

where:

$\Delta F_{\text{p}}$ change in the force with which pressure pushes the tubing upwards: it shortens free tubing or pulls a latched assembly upwards, N
$A_{\text{p}}$ area of the polished bore of the packer (the smooth bore in which the seals sit), taken by the diameter of the seals, mm²
$A_{\text{i}},\ A_{\text{o}}$ areas by the inside and outside diameters of the tubing, mm²
$\Delta p_{\text{i}},\ \Delta p_{\text{o}}$ changes in pressure at the packer in the tubing (which is also the pressure below the packer) and in the annulus, MPa

Units: any consistent units: MPa and mm² give newtons, psi and square inches give pounds-force (lbf)

Holds when

The tubing is open to the reservoir: the pressure in the tubing at the packer equals the pressure below the packer.
The areas are calculated by the diameter of the seals, not of the couplings.
If the packer bore is narrower than the outside diameter of the tubing, $A_{\text{p}} - A_{\text{o}} < 0$ , and the pressure in the annulus also pushes the tubing upwards.

Pressure in the tubing pushes it upwards, and pressure in the annulus pushes it downwards (when the packer bore is wider than the tubing). That is why we hold pressure in the annulus during reservoir treatments: it reduces both this force and the pressure difference across the tubing wall.

Ballooning: the force that would hold the tubing against a change in length

F_{\text{b}} = 2\nu\left(\Delta\bar p_{\text{i}}\,A_{\text{i}} - \Delta\bar p_{\text{o}}\,A_{\text{o}}\right)

where:

$F_{\text{b}}$ ballooning force: positive when the tubing tends to shorten, N
$\nu$ Poisson's ratio of steel, 0.3, —
$\Delta\bar p_{\text{i}},\ \Delta\bar p_{\text{o}}$ changes in pressure in the tubing and in the annulus, averaged over the length of the tubing, MPa

Units: any consistent units (as in the previous formula)

Holds when

Elastic deformation (the tubing springs back when the load goes) of tubing of a single wall thickness.
With fluids of the same density and without friction, the pressure change is the same along the whole length and equals the change at the wellhead.

Pressure inside shortens the tubing, and pressure outside lengthens it. The advanced section shows where the factor $2\nu$ comes from.

In other unit systems (1)

Oilfield units

F_{\text{b}}\,[\text{lbf}] = 0.6\left(\Delta\bar p_{\text{i}}\,A_{\text{i}} - \Delta\bar p_{\text{o}}\,A_{\text{o}}\right)

pressures in psi, areas in square inches.

$0.6 = 2\nu$ at $\nu = 0.3$ .

Temperature: the force on latched tubing and the lengthening of free tubing

F_{\text{t}} = E\,\alpha\,A_s\,\left(-\Delta\bar T\right), \qquad \Delta L_{\text{t}} = \alpha\,L\,\Delta\bar T

where:

$F_{\text{t}}$ thermal force: positive (tension in latched tubing) on cooling, N
$E$ Young's modulus of steel (how stiff it is), 207,000, MPa
$\alpha$ coefficient of linear thermal expansion of steel (how much each unit of length grows per degree), $1.24\cdot10^{-5}$ , 1/°C
$A_s$ cross-sectional area of the tubing wall, mm²
$\Delta\bar T$ change in the length-averaged temperature of the tubing, °C
$L,\ \Delta L_{\text{t}}$ length of the tubing to the packer and its change for free tubing, m

Units: MPa, mm² and °C give the force in newtons; lengths in metres

Holds when

Elastic steel, tubing of a single size.
Heating ( $\Delta\bar T > 0$ ) lengthens free tubing and compresses latched tubing.

The temperature effect is often the largest: every 10 °C changes the length of a kilometre of tubing by 12 cm.

In other unit systems (2)

Metric units

F_{\text{t}}\,[\text{N}] = 2.57\,A_s\,[\text{mm}^2]\left(-\Delta\bar T\,[^\circ\text{C}]\right)

area in mm², temperature in °C.

$2.57 \approx 207{,}000\cdot1.24\cdot10^{-5}$ MPa per °C.

Oilfield units

F_{\text{t}}\,[\text{lbf}] = 207\,A_s\,[\text{in}^2]\left(-\Delta\bar T\,[^\circ\text{F}]\right)

area in square inches, temperature in °F.

$207 = 30\cdot10^{6}$ psi $\cdot\ 6.9\cdot10^{-6}$ 1/°F.

Free and latched seals

\Delta L = \frac{\left(\Delta F_{\text{p}} + F_{\text{b}} + F_{\text{t}}\right)L}{E\,A_s}

where:

$\Delta L$ shortening of free tubing (negative means lengthening); this is how far the seals will move in the packer bore, m
$L$ length of the tubing from the hanger to the packer, m
$E\,A_s$ axial stiffness of the tubing: Young's modulus times wall area, N

Units: SI (forces in newtons, $E$ in pascals, $A_s$ in square metres); it is convenient to take $E$ in MPa and $A_s$ in mm² — their product is also in newtons

Holds when

Vertical tubing of a single size; seal friction is not taken into account.
Helical buckling of free tubing is added separately (advanced section).

If the seals are latched, the tubing pulls the packer upwards with a force $\Delta F_{\text{p}} + F_{\text{b}} + F_{\text{t}}$ , while the axial force in the tubing itself rises by $F_{\text{b}} + F_{\text{t}}$ : the pressure force on the seals is taken directly by the packer.

Worked exampleReservoir treatment through the tubing: seal stroke and load on the packer

Given (assumed data)

Well: 73 × 5.51 mm tubing (62 mm inside) down to a permanent packer at 1950 m in 177.8 × 10.36 mm casing (157.08 mm inside); packer bore 82.55 mm
Treatment: wellhead pressure in the tubing +25 MPa, in the annulus +7 MPa; the tubing and the annulus are filled with the same brine, and before the treatment the pressures at the packer are equal; the average temperature of the tubing falls by 25 °C
Steel: $E$ = 207,000 MPa, $\nu$ = 0.3, $\alpha$ = $1.24\cdot10^{-5}$ 1/°C

Solution

Areas: $A_{\text{o}}$ = 4185.4 (tubing outside), $A_{\text{i}}$ = 3019.1 (tubing inside), $A_s$ = 1166.3 (tubing wall), $A_{\text{p}}$ = 5352.1 (packer bore), and for the casing $A_{\text{c}}$ = 19,379.0, all in mm².
The pressure force on the seals. From below: $(5352.1 - 3019.1)\cdot 25 = 58.33$ kN upwards. From above: $(5352.1 - 4185.4)\cdot 7 = 8.17$ kN downwards. In total, $\Delta F_{\text{p}} = 50.16$ kN upwards.
Ballooning: $F_{\text{b}} = 0.6\cdot(25\cdot 3019.1 - 7\cdot 4185.4) = 27.71$ kN. Cooling: $F_{\text{t}} = 207{,}000\cdot1.24\cdot10^{-5}\cdot 1166.3\cdot 25 = 74.8$ kN. The sum of the three forces is 152.7 kN.
Free seals: the tubing will shorten by $\dfrac{152.7\cdot10^3\cdot 1950}{207{,}000\cdot 1166.3} = 1.23$ m. Of this, 0.41 m comes from the force on the seals, 0.22 m from ballooning and 0.60 m from cooling.
Latched seals: the tubing pulls the packer upwards with a force of 152.7 kN. The axial force in the tubing rises by $27.71 + 74.8 = 102.5$ kN. (The tubing itself, without couplings, weighs about 152 kN in brine.)
The packer itself also holds a pressure difference of $25 - 7 = 18$ MPa across the ring between the casing and the bore: $18\cdot(19{,}379.0 - 5352.1) = 252.5$ kN. In total, the force upwards on the packer is 252.5 kN with free seals and $252.5 + 152.7 = 405.2$ kN with latched seals.

Answer: Free seals need a stroke of at least 1.23 m (with helical buckling, about 1.42 m). Latched seals pass 152.7 kN on to the packer. Together with the pressure difference, the packer must then hold 405.2 kN, which is about 41 tonnes-force.

What it means. Cooling contributed the most: 25 °C shortens 1950 m of tubing by 0.60 m. A short set of seals is not enough here. We need an extended bore or an expansion joint. Latched seals remove the stroke, but they stretch the tubing by 102.5 kN on top of its weight, so we check the tubing and the connections for this load. They also need a packer that holds more than 400 kN from below. A packer set by the weight of the tubing will not hold such a load. Friction during injection lowers the pressure at the packer. A calculation without friction therefore overstates the forces and the stroke (the more so, the higher the rate), which is on the safe side. Helical buckling does the opposite: it adds about 0.19 m to the stroke, and without it the stroke is understated. The data in this example are assumed for illustration.

What changes if…

…we raise the annulus pressure by another 10 MPa. The force on the seals falls by $(A_{\text{p}} - A_{\text{o}})\cdot10$ ≈ 11.7 kN, and ballooning falls by $0.6\cdot10\cdot A_{\text{o}}$ ≈ 25.1 kN. The pressure difference across the packer itself falls by 10 MPa, which means 140 kN less load. This is the main way to take load off the tubing and the packer during a treatment. The limit is the strength of the production casing.
…we inject heated fluid. The largest contribution, the thermal one, gets smaller.
…we use a packer with a bore narrower than the outside diameter of the tubing. The annulus pressure then pushes the tubing upwards, not downwards.
…we switch the well to producing hot fluid. The tubing gets longer. With a latched assembly, the tubing is put in compression, and there is a risk of buckling (bending). With a locator assembly, the locator lands on the packer and passes a compressive force to it.

CalculatedCalculated shortening of the free tubing of the example under the same treatment (+25 MPa in the tubing, cooling by 25 °C), against the pressure held in the annulus. One curve is without helical buckling, and one is with it. The dashed line is the seal stroke of 1.5 m. The pressure in the annulus reduces the force on the seals, ballooning and buckling.

Seal assemblies, expansion joints and disconnect devices

The way the tubing is connected to a permanent packer decides whether a change in length turns into a force. A seal assembly is a short piece of pipe (a sub) with a set of seals. It sits on the lower end of the tubing and enters the polished bore of the packer.

Anchor seal assembly (latch): It snaps into the packer, and we usually release it by turning the tubing. The tubing does not move, so every change in length turns into a force. There is tension on cooling and with pressure in the tubing, and compression on heating. The seals do not rub against the bore. This is the best option for gas and for large pressure differences, if the tubing and the packer can withstand the forces.
Locator seal assembly: The seals move freely in the packer bore. A locator shoulder at the top stops the tubing from going down below its set position. The seals must be long enough to cover the stroke upwards. The tubing also needs room to get longer, so the shoulder is left above the packer with a gap. Otherwise the compression would act on the packer. The packer bore is extended by a pipe below it, called a seal bore extension.
Expansion joint (telescopic joint, travel joint): Two concentric pipes (one inside the other) with seals between them. The stroke is several metres. We install it above the packer when the expected movement is larger than a locator assembly allows, for example in high-rate wells with large temperature changes.
Polished bore receptacle (PBR) at the top of a liner: The seal assembly of the tubing enters the receptacle (a smooth sleeve) directly, and the receptacle takes the place of the packer: it isolates the annulus and lets the tubing move. It gives a large bore. This is how we build completions of almost constant diameter from the wellhead to the reservoir.
On-off tool (disconnect device): It lets us pull the tubing out while the packer and the pipe below it stay in place. Before that, we close off the reservoir with a plug in a nipple below the disconnect device. We install the components that we change more often, the safety valve and the pump, above the disconnect device.

What to choose according to the expected change in length

Movement: small; sealing matters most
Solution: anchor seal assembly
Trade-off: forces in the tubing and on the packer, a risk of buckling on heating

Movement: up to a metre or two
Solution: locator seal assembly with an extended bore
Trade-off: friction and wear of the seals, exact space-out (fitting the length of the string)

Movement: several metres
Solution: expansion joint
Trade-off: one more moving seal in the string

Movement	Solution	Trade-off
small; sealing matters most	anchor seal assembly	forces in the tubing and on the packer, a risk of buckling on heating
up to a metre or two	locator seal assembly with an extended bore	friction and wear of the seals, exact space-out (fitting the length of the string)
several metres	expansion joint	one more moving seal in the string

The boundaries are approximate: the decision is made by calculation for all operations — production, shut-in, injection, reservoir treatment, pressure testing.

Subsurface safety valves

The valves of the Christmas tree are the main way to shut in a well. But the tree is removed before a workover, its valves need servicing, and it can lose its seal or be damaged by fire, by machinery or, offshore, by the impact of a ship. For this case we install a subsurface safety valve (SSSV; the Russian-language standards call it a “downhole safety valve”) in the tubing below the wellhead. Offshore, in wells with hydrogen sulfide, and near populated areas, it is, as a rule, mandatory.

Surface-controlled (SCSSV): It is held open by the pressure in a thin hydraulic control line clamped to the outside of the tubing. When the pressure in the line is bled off, a spring closes the valve. The pressure is bled off on a signal from the emergency system, if the line is cut, or if the wellhead is damaged. A failure of the control system shuts the well in rather than leaving it open. Most valves are designed this way.
Direct-controlled (“storm choke”): It closes by itself when the flow through it exceeds the design rate or the pressure falls below a set value. There is no control line. It works only in a strong blowout (an uncontrolled release of fluid from the well), and we have to change its setting whenever the operating mode of the well changes. This is why surface-controlled valves have replaced it.

The closing element is a hinged flapper or a ball that turns through 90°. The control line pressure acts on a piston and pushes down a moving sleeve. The sleeve opens the flapper or turns the ball, and a spring returns the sleeve. Flappers are simpler and more widespread. A closed valve holds pressure from below, and we cannot open it against a large pressure difference. First we equalise the pressure above it, either by pumping into the tubing or with a built-in equalising valve. The built-in valve is more convenient, but it adds seals and fails more often.

Tubing-retrievable and wireline-retrievable valves

Feature: Bore
Part of the tubing string (TRSV): almost the same as the tubing
Set on wireline in a nipple (WRSV): noticeably smaller: it limits the rate and access below

Feature: Replacement after a failure
Part of the tubing string (TRSV): only by pulling the tubing; temporarily, an insert valve run on wireline inside the failed one
Set on wireline in a nipple (WRSV): by wireline operations, without pulling the tubing

Feature: Where it is appropriate
Part of the tubing string (TRSV): high-rate wells, especially gas wells
Set on wireline in a nipple (WRSV): where the bore does not limit the rate and quick replacement matters

Feature	Part of the tubing string (TRSV)	Set on wireline in a nipple (WRSV)
Bore	almost the same as the tubing	noticeably smaller: it limits the rate and access below
Replacement after a failure	only by pulling the tubing; temporarily, an insert valve run on wireline inside the failed one	by wireline operations, without pulling the tubing
Where it is appropriate	high-rate wells, especially gas wells	where the bore does not limit the rate and quick replacement matters

We set the valve below the zone where hydrates (ice-like solids that form from gas and water) may form in the tubing or wax may be deposited, because they would stop the valve from closing. We also set it below the possible crater in an open blowout, and offshore, below the seabed. But we cannot set it too deep either. The control line gets longer, the opening pressure gets higher and, above all, the column of fluid in the line presses on the piston even when the pressure at the surface has been bled off.

Control line pressure and the maximum setting depth of a safety valve

p_{\text{c}} \ge p_{\text{i}}(H) + p_{\text{open}} - \rho_{\text{c}}\,g\,H, \qquad H_{\max} = \frac{p_{\text{close}}}{\rho_{\text{c}}\,g}

where:

$p_{\text{c}}$ pressure in the control line at the surface (gauge), Pa
$p_{\text{i}}(H)$ pressure in the tubing at the depth of the valve; the highest is in the shut-in well, Pa
$p_{\text{open}}$ full-opening pressure from the data sheet: by how much the pressure in the line at the valve must exceed the pressure in the tubing, Pa
$p_{\text{close}}$ closing pressure from the data sheet: the excess of the line pressure over the tubing pressure at which the spring begins to close the valve, Pa
$\rho_{\text{c}}$ density of the fluid in the control line, kg/m³
$H$ vertical setting depth of the valve, m

Units: SI (pressures in pascals, $\rho_{\text{c}}$ in kg/m³, $H$ in metres)

Holds when

The valve has a piston. The tubing pressure acts on it from below, and the line pressure acts on it from above. Friction and areas are already included in the data-sheet opening and closing pressures.
$H_{\max}$ has no safety margin. With the line bled off and zero pressure in the tubing, the column of fluid in the line must not exceed the closing pressure. The manufacturer and the design add a margin.
For greater depths we use valves with a balance line or a gas chamber. The column of fluid in the line has almost no effect on them.

Pressure in the tubing helps to close the valve and makes it harder to open, while the column of fluid in the line does the opposite. That is why we hold the line at a pressure above the tubing pressure of the shut-in well, and the spring limits the setting depth.

In other unit systems (2)

Metric units

p_{\text{c}} \ge p_{\text{i}}(H) + p_{\text{open}} - 0.00981\,\rho_{\text{c}}H, \qquad H_{\max} = \frac{p_{\text{close}}}{0.00981\,\rho_{\text{c}}}

pressures in MPa, $\rho_{\text{c}}$ in g/cm³, $H$ in m.

0.00981 MPa is the pressure of a column of fluid with a density of 1 g/cm³ and a height of 1 m.

Oilfield units

p_{\text{c}} \ge p_{\text{i}}(H) + p_{\text{open}} - 0.052\,\rho_{\text{c}}H, \qquad H_{\max} = \frac{p_{\text{close}}}{0.052\,\rho_{\text{c}}}

pressures in psi, $\rho_{\text{c}}$ in ppg, $H$ in feet.

0.052 psi/ft per ppg — from the oilfield formula for hydrostatic pressure.

Worked exampleControl line pressure and the setting depth of a safety valve

Given (assumed data)

Valve: at 250 m; closing pressure 2.6 MPa, full-opening pressure 4.0 MPa (from the data sheet)
Control line: hydraulic oil, 0.87 g/cm³
Tubing pressure at the valve in the shut-in well: 20.0 MPa

Solution

The column of oil in the line: $0.00981\cdot0.87\cdot250 = 2.13$ MPa.
Will the valve close with the line bled off and zero pressure in the tubing? 2.13 < 2.6 MPa — yes, with a margin of 0.47 MPa. The maximum depth: $H_{\max} = 2.6/(0.00981\cdot0.87) = 305$ m.
The line pressure at the surface needed to keep the valve fully open even in the shut-in well: $20.0 + 4.0 - 2.13 = 21.87$ MPa.
The same valve at 400 m: the column of oil is $0.00981\cdot0.87\cdot400 = 3.41$ MPa — more than the closing pressure.

Answer: The line must be held at no less than 21.87 MPa; a valve with such a spring can be set no deeper than 305 m (without a margin).

What it means. The line pressure must be higher than the pressure of the shut-in well. So the line, its connections and the pump of the control unit are rated for high pressure. At 400 m the column of oil (3.41 MPa) would be higher than the closing pressure. At a low tubing pressure the valve could then stay open just when it is needed. For such a depth we use a valve with a stronger spring, or one that compensates for the column of fluid in the line. The data in this example are assumed for illustration.

We test the safety valve from time to time. We close it, bleed off the pressure above it and watch how the pressure builds up. The standard allows a small leak in such a test: for gas, about 0.42 m³/min (15 standard cubic feet per minute). Many companies require stricter limits. ISO 10432 (GOST ISO 10432-2014) sets the requirements for the valves themselves. ISO 10417 (GOST ISO 10417-2014) sets them for the design, installation, operation and testing of the systems.

Landing nipples and what is installed in them

A landing nipple is a short piece of pipe (a sub) in the tubing string. Inside it has a profile for a lock and a polished sealing bore. We run a lock mandrel, with locking keys and seals, into the nipple on wireline or coiled tubing (a long, continuous pipe that is unreeled into the well). The lock carries the device that we need:

a blanking plug (a solid plug that blocks the tubing) — to pressure-test the tubing, to set a hydraulic packer, to close off the reservoir before a workover or before disconnecting the tubing
a check valve that holds pressure from above and lets the flow pass upwards
a downhole pressure gauge and thermometer for surveying the well
a downhole choke (a restriction in the tubing) — to reduce the pressure higher up the wellbore, for example against hydrates
a wireline-retrievable safety valve — into a nipple of special design to which the control line is connected

Nipples can be selective or no-go. Selective nipples are several identical ones in a single string. The lock lands in the right one because of the way the running tool is adjusted, or because of the profile of the keys. A no-go nipple has a slightly smaller bore at the bottom, and the lock cannot pass further. A no-go nipple is usually the lowest one. It also stops a device if the device is dropped higher up. Every nipple narrows the bore. So from top to bottom the sizes are either the same (selective) or get smaller, and a device for the lowest nipple must pass through all the upper ones.

Restrictions create eddies that wear away (erode) the tubing above and below the nipples and the safety valve. We install flow couplings there. They have extra wall thickness to allow for wear. If a gauge in the lowest nipple blocks the end of the tubing, the flow enters higher up through a perforated joint (a short pipe with holes in its wall). We fit the lower end of the string with a wireline entry guide. A tool that was run below the tubing returns into the pipe through the guide without catching on the end face. The lesson Wireline operations covers how devices are run and retrieved on wireline.

Sliding sleeves and side-pocket mandrels

A sliding sleeve is a short pipe (a sub) with ports (windows) in its wall. Inside it moves a sleeve that has its own ports, with seals above and below them. A tool on wireline or coiled tubing shifts the sleeve by jarring up or down (hitting it with a sharp upward or downward blow). The ports line up (the sleeve is open) or move apart (it is closed). Through the sleeve we flush the well and kill it by circulation (killing before a workover), we replace the fluid in the tubing or the annulus, and in wells with several reservoirs we open access to the upper reservoir.

Before opening the sleeve, we equalise the pressures in the tubing and the annulus. With a large pressure difference we cannot shift the sleeve, and a sudden flow through the ports will damage the seals and may throw the tool up the hole.
Sand, deposits, or poor transmission of the jarring blows in a deviated hole may stop us closing the sleeve. Then we set an isolation sleeve inside it, which covers the ports. This narrows the bore.
If there is no sliding sleeve, or it does not open, we make a hole in the tubing with a mechanical puncher or a small shaped charge (an explosive charge that cuts a hole) that does not damage the casing behind the tubing (Shaped-charge perforating). We cannot close such a hole, so the tubing is replaced afterwards.

A side-pocket mandrel is a short pipe (a sub) with a pocket set off to one side of the main bore, and with ports from the pocket into the annulus. The main bore stays full-bore (it is not narrowed). We set valves in the pocket, and retrieve them from it, on wireline, with a tool that is deflected towards the pocket (a kickover tool):

gas lift valves — pass gas from the annulus into the tubing (Gas lift: principles and choice of operating conditions; the mechanics of the valves and unloading the well are in Unloading a gas lift well: valves and mandrels)
chemical injection valves — for corrosion and scale inhibitors and pour-point depressants (chemicals that keep the oil fluid in the cold); they open when the annulus pressure exceeds a set value
blanking plugs — the mandrels are built into the string in advance, and gas lift is started years later by replacing the plugs with valves without pulling the tubing
circulating valves with a shear element (a pin that breaks at a set load) — they open a path between the tubing and the annulus for killing the well; the path can be closed again only by replacing the valve on wireline

Terms and symbols

Retrievable and permanent packers: The first kind is released by pulling or turning and is pulled out with the tubing. The second kind is set on a separate run and removed by milling.
Seal assembly: A set of seals on the lower end of the tubing, in the bore of a permanent packer. It can be an anchor assembly, a locator assembly, or one paired with an expansion joint.
Tubing movement: The lengthening or shortening of the tubing because of the pressure force on the seals, ballooning, temperature and helical buckling.
Subsurface safety valve: A valve in the tubing below the wellhead. A surface-controlled valve is closed by a spring when the pressure in the control line is bled off.
Landing nipple and lock mandrel: A short pipe with a profile and a sealing bore, and the mandrel with a device that is run into it on wireline.
Sliding sleeve: A sleeve with ports that connects the tubing with the annulus.
Side-pocket mandrel: A short pipe with a side pocket for a gas lift, chemical injection or circulating valve that can be changed on wireline.
Tubing hanger: The component in the tubing head on which the tubing hangs and which seals the annulus at the wellhead.

Common mistakes and misconceptions

Wrong

“The packer is set, so the tubing goes nowhere.”

In fact

Pressure and temperature change the length of the tubing by a metre or more. If the seals are latched, this turns into a force. In the example in this lesson, the tubing pulls the packer upwards with a force of about 150 kN. With the pressure difference as well, the packer has to take more than 400 kN.

Wrong

“A mechanical packer holds pressure from above and from below equally well.”

In fact

A packer set by the weight of the tubing holds a pressure difference from above, while pressure from below and the shortening of the tubing take the load off it. For loads in both directions, we need slips in both directions, a hydraulic hold-down or a permanent packer.

Wrong

“It is safer to set the safety valve deeper.”

In fact

The deeper the valve, the harder the column of fluid in the control line presses, and the spring may not close the valve when the line is bled off. The closing pressure limits the depth. In the example in this lesson the limit is about 300 m.

Wrong

“Pressure in the annulus during a treatment is an extra load on the casing and should be bled off.”

In fact

It reduces the pressure difference across the tubing wall, the force on the seals and ballooning. In the example in this lesson, every 10 MPa in the annulus reduces the force with which latched tubing pulls the packer by about 37 kN. It reduces the load from the pressure difference across the packer itself by another 140 kN. The strength of the production casing and the packer sets the limit.

Wrong

“You can never have too many nipples — let's put in a few extra just in case.”

In fact

Every nipple narrows the bore and creates eddies and erosion, and a device for the lowest one must pass through all the upper ones. We install as many nipples as the work plan needs for plugs, gauges and valves over the whole service life of the well.

In depth: advanced detail

Helical buckling of free tubing

Why does the pressure in the tubing bend it, even though it acts from the inside? The pressure force on the seals, together with the difference between the pressures inside and outside the tubing, acts on the stability of the tubing as a compressive force applied to the bottom of the free string. It is called the fictitious (effective) force. This force matters for buckling, not the true axial force in the steel (true axial force). Tubing that hangs freely in a liquid with no pressure difference stays straight. When the pressure inside is higher than outside, the lower part of free tubing bends into a helix (a spiral) and lies against the walls of the casing.

Shortening of free tubing due to helical buckling

F_{\text{f}} = A_{\text{p}}\,(p_{\text{i}} - p_{\text{o}}), \qquad n = \frac{F_{\text{f}}}{w}, \qquad \Delta L_{\text{buck}} = \frac{r^2 F_{\text{f}}^2}{8\,E\,I\,w}

where:

$F_{\text{f}}$ fictitious force at the packer; buckling occurs if it is positive, N
$p_{\text{i}},\ p_{\text{o}}$ pressures in the tubing and in the annulus at the packer, Pa
$n$ length of the buckled section above the packer (up to the neutral point, where the buckling force falls to zero), m
$w$ weight per metre of tubing in the fluid: steel plus the fluid inside minus the fluid displaced outside, N/m
$r$ radial clearance between the tubing and the casing, m
$I$ second moment of area of the tubing cross-section (how well it resists bending), $\pi(D^4 - d^4)/64$ , m⁴
$\Delta L_{\text{buck}}$ shortening of the tubing due to buckling, m

Units: SI

Holds when

The well is vertical, the tubing is of a single size, the seals are free, and there is no friction against the casing.
The formula holds while $n \le L$ . If the whole string is buckled, we need a different formula.
With latched tubing, the tension that the packer transmits reduces the fictitious force, and there may be no buckling.

A helix (spiral) whose pitch (the distance between turns) is set by the force $F$ shortens each metre of tubing by $r^2F/(4EI)$ . The force falls from $F_{\text{f}}$ at the packer to zero at the neutral point. Adding up (integrating) over the buckled section gives the formula. The shortening grows with the square of the pressure difference at the packer and with the square of the clearance. In wide casing, thin tubing buckles more.

Worked exampleBuckling in the reservoir treatment example

Given (assumed data)

Pressure difference at the packer: the pressures are equal before the treatment; during the treatment $25 - 7 = 18$ MPa
Tubing and brine: mass of a metre of steel 9.17 kg/m; brine 1.05 g/cm³ inside and outside

Solution

Fictitious force: $F_{\text{f}} = 5352.1\cdot18 = 96.34$ kN.
Weight of a metre in brine: steel $9.17\cdot9.81 = 89.96$ N/m minus the brine in the volume of the wall $1050\cdot9.81\cdot1166.3\cdot10^{-6} = 12.01$ N/m, that is, $w = 77.95$ N/m. The buckled section is $n = 96{,}338/77.95 = 1236$ m. This is less than 1950 m, so the formula applies.
Clearance $r = (157.08 - 73)/2 = 42.04$ mm, $I = 6.6866\cdot10^{-7}$ m⁴.
Denominator: $8EIw = 8\cdot 2.07\cdot10^{11}\cdot 6.6866\cdot10^{-7}\cdot 77.95 = 8.631\cdot10^{7}$ N²·m. Shortening: $\Delta L_{\text{buck}} = 0.04204^2\cdot 96{,}338^2/(8.631\cdot10^{7}) = 0.190$ m.

Answer: Buckling adds 0.190 m, and the total seal stroke is about 1.42 m.

What it means. Buckling gives about 13% of the total stroke. At an annulus pressure of 0 MPa it would almost double. At 18 MPa it would fall to about 3 cm. At 25 MPa, when the pressure difference at the packer disappears, it would vanish altogether (see the graph in the main part of the lesson). The data in this example are assumed for illustration.

Where the factor 2ν in the ballooning formula comes from

In a thick-walled pipe under pressure $p_{\text{i}}$ inside and $p_{\text{o}}$ outside, the radial stress (across the wall) and the hoop stress (around the pipe) change across the wall thickness (the Lamé solution). But their sum is constant: $\sigma_r + \sigma_\theta = 2\,(p_{\text{i}}A_{\text{i}} - p_{\text{o}}A_{\text{o}})/A_s$ . By Hooke's law, the axial strain (the relative change in length) is $\varepsilon_z = \left[\sigma_z - \nu(\sigma_r + \sigma_\theta)\right]/E$ .

A pipe with no additional axial force shortens by $\nu(\sigma_r + \sigma_\theta)/E$ per metre. This is ballooning. If the tubing is latched, then $\varepsilon_z = 0$ , and an axial stress $\sigma_z = \nu(\sigma_r + \sigma_\theta)$ appears in the steel. The force is $\sigma_z A_s = 2\nu(p_{\text{i}}A_{\text{i}} - p_{\text{o}}A_{\text{o}})$ . Written with the changes in pressure, this gives the formula of the lesson. At $\nu = 0.3$ the factor is 0.6. Pressure acts on the walls along the whole length. This is why the formula contains length-averaged pressure changes.

Connections to other topics

Well completion: opening the reservoir and the tubing string — the functions of the tubing string, killing by circulation and well barriers.
Casing design — loads on the production casing when the tubing leaks, and the true axial force in the pipe.
Casing: purpose, types, properties and standards — pipe sizes, strength and connections — for tubing they are the same as for casing.
Gas lift: principles and choice of operating conditions — what side-pocket mandrels are for and how gas lift works.
Shaped-charge perforating — how the reservoir is opened and what is used to punch a hole in the tubing if a sliding sleeve does not open.
Multiple-zone completions: commingled and segregated production — arrangements with several packers and sliding sleeves for several reservoirs.
Wireline operations — how plugs, valves and gauges are set and retrieved on wireline.

Check yourself

1
Cold water is being injected into an injection well, and the pressure below the packer is higher than above it. Which retrievable packer is the most suitable?
1. A.An inflatable packer
2. B.Set by tubing weight (compression-set), without a hold-down
3. C.Tension-set, or with slips in both directions
4. D.Any: a packer holds a pressure difference equally well in both directions
Show the answer
Correct answer: C. Tension-set, or with slips in both directions
Pressure from below and the shortening of the tubing as it cools pull the tubing and the packer upwards. A tension-set packer only presses its slips harder as a result, while a packer set by the weight of the tubing is unloaded and may move. If the pressure difference can change direction, we need slips in both directions or a hydraulic hold-down.
2
In the example of the lesson, the annulus pressure during the treatment was raised from 7 to 17 MPa. By how much will the pressure force on the seals and the ballooning force decrease? Areas: $A_{\text{o}}$ = 4185.4 mm², $A_{\text{p}}$ = 5352.1 mm².
Show the answer
The force on the seals will fall by $(5352.1 - 4185.4)\cdot10 = 11{,}667$ N ≈ 11.7 kN, and ballooning by $0.6\cdot10\cdot4185.4 = 25{,}112$ N ≈ 25.1 kN. That is about 37 kN in all. The stroke of free tubing, leaving buckling aside, will fall by $36.8\cdot10^3\cdot1950/(207{,}000\cdot1166.3) \approx 0.30$ m. Buckling will be smaller too, because the pressure difference at the packer will fall from 18 to 8 MPa. The price is a higher pressure on the production casing.
3
A gas well with a rate close to the flow capacity of the tubing. Which safety valve should be chosen?
1. A.No valve is needed if there are two valves at the wellhead
2. B.Part of the tubing string: its bore is almost as large as the bore of the tubing
3. C.A direct-controlled valve with no control line
4. D.Wireline-retrievable: it is easier to replace
Show the answer
Correct answer: B. Part of the tubing string: its bore is almost as large as the bore of the tubing
A valve set on wireline in a nipple narrows the bore noticeably and would limit the rate of a gas well. A valve that is part of the tubing string is almost full-bore; if it fails, an insert valve can be set inside it temporarily on wireline.
4
During injection the tubing will shorten by 1.0 m, and during the production of hot oil it will lengthen by 0.4 m. A locator seal assembly is used, with a seal length of 1.6 m. How should the assembly be spaced out, and will the seals be long enough?
Show the answer
The tubing cannot go lower than the position in which the locator sits on the packer. So we leave the locator above the packer with a gap of at least 0.4 m (better with a margin). Otherwise the lengthening would turn into a compressive force on the packer. The seals must stay in the bore when the tubing rises by 1.0 m from its position at setting. That is 1.0 + 0.4 = 1.4 m from the lowest position (locator on the packer), so 1.6 m is enough with a small margin. If it were not enough, we would install a seal bore extension below the packer, or an expansion joint.
5
A sliding sleeve was opened for killing but could not be closed after the workover. What is usually done without pulling the tubing?
1. A.Set an isolation sleeve inside to cover the ports
2. B.Pump cement into the annulus
3. C.Punch a second hole next to it
4. D.Nothing: an open sleeve does not interfere with production
Show the answer
Correct answer: A. Set an isolation sleeve inside to cover the ports
We set an isolation sleeve on wireline in the sealing bores above and below the ports. This cuts off the path between the tubing and the annulus, but it narrows the bore. We cannot leave an open sleeve: the production would go into the annulus, and the casing would be exposed to its pressure.
6
Why can a safety valve with a spring designed for a closing pressure of 2.6 MPa not be set at 500 m when the control line contains oil of 0.87 g/cm³?
Show the answer
A column of oil 500 m high creates $0.00981\cdot0.87\cdot500 \approx 4.3$ MPa at the valve even when the pressure at the surface has been bled off. This is more than the closing pressure: at a low tubing pressure the spring will not move the sleeve, and the valve will stay open. We need a valve with a stronger spring, a valve with a balance line, or a setting no deeper than about 300 m.

Updated: 15 October 2026