Wellheads and running casing

Quick refresher

The casing wellhead is a stack of flanged units: the lower casing head on the surface casing, a spool for each following string, then the tubing head and the Christmas tree. Each string hangs on a hanger in the bowl (the conical seat) of its head. A seal closes the annulus between strings, and the crew can reach it through side outlets.
A mandrel (threaded) hanger is screwed onto the top pipe, so the length of the string must be adjusted exactly (this is called spacing out). A slip hanger grips the pipe at any point, so the string can be hung at any depth and with the tension that is needed.
We choose the pressure class (2000…20,000 psi, in GOST 30196 14…140 MPa) from the highest expected wellhead pressure. A spool with a seal on the string can have an upper flange one step higher than the lower one. This is how the crew moves up to a higher class.
The crew screws the thread together under control. For a buttress thread (a trapezoidal thread), the control is the position of the end of the coupling relative to the triangle mark. For a round thread (a triangular thread), it is the torque (the turning force). For premium connections (gas-tight connections), it is a torque–turn plot in which the stop faces (shoulders) meet.
The crew fills a string that has a float valve (a one-way valve at the bottom of the string). Otherwise the string floats (in the example of the lesson an unfilled string weighs about a quarter of its full weight on the hook). And if the valve fails, the level in the annulus drops and the pressure on the formations falls.
While running, the string pushes mud out through the narrow annulus, and the pressure on the walls rises more than when drill pipe is run. We choose the running speed and the fill-up so that we stay in the window between the pore pressure (the pressure of the fluid in the rock) and the leak-off pressure (the pressure at which the rock starts to take mud).

Key relationships

w = g\left[m\left(1 - \dfrac{\rho_{\text{out}}}{\rho_{\text{steel}}}\right) + (\rho_{\text{in}} - \rho_{\text{out}})\,A_{\text{in}}\right]

weight of a metre of string when the fluids inside and outside differ (an empty string has $\rho_{\text{in}} = 0$ )

y = L_{\text{empty}}\,\dfrac{A_{\text{in}}}{A_{\text{in}} + A_{\text{ann}}}

how far the level in the annulus drops if the float valve fails on a string whose upper part, of length $L_{\text{empty}}$ , is unfilled

v_{\text{ann}} = v_{\text{pipe}}\left(\dfrac{d^2}{D^2 - d^2} + K_c\right)

speed of the mud in the annulus while a string with a closed bottom is run

p_{\text{wh}} = p_{\text{shoe}} - 0.00981\,\rho_{\text{g}}\,h_{\text{shoe}}

the highest expected wellhead pressure of a closed well full of gas (MPa, g/cm³, m)

Common pitfalls

Choosing the class of the lower casing head only by the pressure of the next interval. The spool above it usually raises the class by only one step.
Taking off the protector caps and applying thread compound in advance, on the catwalk (the walkway where the pipes lie before they are lifted). Dirt gets into the compound.
Assuming that, because the float valve holds, the fill-up can wait until the end of the run.
Running a string at the same speed as drill pipe. At the same speed, the pressure surges of a string are several times larger.
Testing the seals at the wellhead with a pressure equal to the flange class, without comparing it with the collapse pressure of the string.

After this lesson you can

describe the casing wellhead (the lower casing head, spools, hangers, seals and side outlets), and explain how it holds the weight of the strings, seals off the spaces between them and is the base for the preventers
choose the pressure class of the casing heads from the highest expected wellhead pressure, explain how a seal on the string allows a step up to a higher class, and name the factory and field test pressures
list the preparation for running and the rig equipment for casing: drifting (checking each pipe with a gauge), the tally (the list of pipe lengths), elevators (clamps that lift a pipe), the spider (the clamp that holds the string at the rig floor), tongs (wrenches that screw the pipes together) and fill-up devices
explain how the screwing together of connections is controlled (by the position of the coupling, by torque, by the torque–turn plot), and why thread compound (thread grease) is needed
calculate the fill-up and the hook load of a string (the weight that the hook of the hoisting system carries), allowing for buoyancy and for different fluids inside and outside; estimate the pressure surges while running; choose the fill-up routine and the running speed
describe landing the string, cutting it off, installing the next spool and testing the seals

The casing wellhead: what it is made of

On the wellhead of a well on land or on an offshore platform stands the casing wellhead, a stack of steel flanged units. The crew builds it from the bottom up as it runs the casing strings. Its jobs are these:

hold the weight of each string, hundreds of tonnes, and pass it to the surface casing and the conductor, and through them to the cement sheath and the rock
close the space between neighbouring strings (the annulus between strings) and give access to it through side outlets with valves: through them the crew watches the pressure and, if needed, pumps fluid in or lets it out
be the base for the blowout preventers (BOP) while the next interval is drilled (Blowout prevention equipment), and after completion for the tubing head and the Christmas tree
hold the pressure of a closed well together with the string: the pressure shell holds only as much as its weakest link

Lower casing head (casing head, casing head housing): Single-flanged. At the bottom, the crew screws it onto the thread of the surface casing, or slides it over the pipe and welds it. At the top is a flange for the preventers. Inside is a seat, the bowl, with a conical surface for the hanger of the next string. On the side are outlets.
Casing spool: A double-flanged casing head. Its lower flange sits on the previous head, and the preventers for the next interval go on its upper flange. In the lower flange is a seal on the pipe of the previous string, and inside is a bowl for the hanger of the next one. There are as many spools as there are strings hanging in the stack above the surface casing.
Tubing head: The last spool. The tubing, the pipe that carries the production, hangs in it on a tubing hanger, and the Christmas tree sits on top. This is already completion equipment (the tubing string).

SchematicThe casing wellhead of the lesson's well after the 244.5 mm string is run. The lower casing head stands on the 339.7 mm surface casing. The 244.5 mm string hangs in its bowl on a slip hanger with a seal. Above it is a crossover spool with a secondary seal on the cut-off end of the string and a test port between the seals. The crew puts the preventers for drilling the next interval on top of the spool. Not to scale.

The flanges are joined with studs (threaded bolts) and sealed with steel ring gaskets (in GOST 28919, of octagonal section). A gasket lies in grooves in both flanges and is squeezed as the studs are tightened. For high-pressure gaskets, the well pressure also presses the gasket against the walls of the grooves. The size of a flange sets the nominal bore (the size of the opening through it): in the inch series 7 1/16, 9, 11, 13⅝, 16¾, 21¼″…, and in GOST 30196 it is 180, 230, 280, 350, 425, 540 mm… The pressure class sets the thickness of the flange and the number and diameter of the studs. The bore of each head must be larger than the drift diameter of the string under it. Otherwise the bit for the next interval will not pass.

While the crew drills through the head, a wear bushing (a protective sleeve) covers the bowl. The tool joints of the drill string (the thick threaded ends of the pipes) and the bit then do not scratch the conical surface on which the hanger will later sit. Before running the next string, the crew pulls the bushing out.

A wellhead made of separate spools is simple and flexible: the crew can put a spool on for any new string. But for each spool it has to take the preventers off and assemble them again, and each flanged joint is one more place for a leak. So compact (multibowl) casing heads (also called speed heads) are common too. One body with several bowls takes two or three strings and the tubing hanger, and the crew lowers the hangers through the preventers, without removing them. This saves hours of rig time and the height under the rotary table (the turntable in the rig floor). But the casing programme must be known in advance: a string of an unplanned size cannot be hung in such a head. Under water the wellhead is built differently, with a subsea wellhead on the seabed (Offshore and subsea drilling).

Hangers and seals

The crew hangs a string that has reached its planned depth in the bowl of the casing head, on a hanger. The hanger passes the weight of the string to the conical landing surface of the bowl. A seal in the hanger or above it (the pack-off) closes the annulus between strings.

Two kinds of casing hangers

Hanger: Mandrel (threaded) hanger
How it is built: a thick-walled short pipe (a sub) with a conical outside surface for landing is screwed onto the top pipe; it sits in the bowl together with the string, and the seals are on its body
When it is used: the string reaches its planned depth freely; compact heads, where the hanger is lowered through the preventers

Hanger: Slip hanger
How it is built: a split body with slips (toothed wedges) grips the pipe and is lowered into the bowl; under the weight of the string the slips bite into the pipe; the seal is pressed tight by the weight or by bolts
When it is used: the string has not reached the bottom, or it must be hung in tension; after cementing, when the string is no longer moved

Hanger	How it is built	When it is used
Mandrel (threaded) hanger	a thick-walled short pipe (a sub) with a conical outside surface for landing is screwed onto the top pipe; it sits in the bowl together with the string, and the seals are on its body	the string reaches its planned depth freely; compact heads, where the hanger is lowered through the preventers
Slip hanger	a split body with slips (toothed wedges) grips the pipe and is lowered into the bowl; under the weight of the string the slips bite into the pipe; the seal is pressed tight by the weight or by bolts	the string has not reached the bottom, or it must be hung in tension; after cementing, when the string is no longer moved

In the code of a casing head in GOST 30196, a letter marks the type of the hanger: K for a slip hanger and R for a threaded one (letters transliterated from the Russian words for “slip” and “threaded”).

SchematicA mandrel hanger and a slip hanger in the bowl of a casing head (half sections). The mandrel hanger is screwed onto the top pipe and sits with its taper on the taper of the bowl. The slip hanger grips the pipe: the slips slide down the taper of the bowl and bite into the pipe under its weight. Not to scale.

A mandrel hanger requires spacing out the string: the shoe (the bottom end of the string) must reach the right depth exactly when the hanger lands in the bowl. The crew picks the pipes to get the right length from the tally, the list of pipes with their measured lengths, and makes the last correction with short pipes called pup joints (parts of a string). If the string stops on a ledge above the planned depth, the mandrel hanger cannot be landed. So the crew keeps a slip hanger ready for this case.

There are two seals at the wellhead. The primary seal is the hanger seal (the pack-off; in Russian equipment, the packer of the casing head). It closes the annulus under it. The secondary seal sits in the lower flange of the next spool. It grips the cut-off end of the string. It separates the inside of the spool, and so the well pressure while the next interval is drilled, from the flanged joint below. Between the seals is a cavity with a test port. Through it the crew pressure-tests the flanged joint and both seals, without loading the string from the inside or the rock.

The secondary seal allows a higher pressure class. Inside the spool, the well pressure acts only on the area limited by the seal on the string. This is much smaller than the area of the ring gasket of the lower flange, so the lower joint stays within its class. So a crossover spool can have an upper flange of a higher class than the lower one, usually by one step: for example, 35 MPa at the bottom and 70 MPa at the top. In GOST 30196 such a head is marked, for example, GK2K-350×35-280×70. This means double-flanged, with a slip hanger, a lower flange with a nominal bore of 350 mm for 35 MPa, and an upper flange of 280 mm for 70 MPa. A single-flanged lower head is marked GK1K, followed by the diameter of the string it sits on and the diameter of the string it hangs. The letters are transliterated from the Russian.

Pressure class and tests

We choose the pressure class (rated working pressure) of a casing head in the same way as the working pressure of preventers (the working pressure of BOP equipment). It comes from the highest expected wellhead pressure while the interval is drilled with preventers on this head. The design case is a closed well full of gas. The pressure at the shoe cannot then be higher than the fracture pressure of the rock under it, the pressure at which the rock breaks (the same case for the string).

Pressure classes of wellhead equipment

Class, psi: 2,000
MPa: 13.8
GOST series, MPa: 14
Body test, psi: 4,000

Class, psi: 3,000
MPa: 20.7
GOST series, MPa: 21
Body test, psi: 6,000

Class, psi: 5,000
MPa: 34.5
GOST series, MPa: 35
Body test, psi: 7,500

Class, psi: 10,000
MPa: 68.9
GOST series, MPa: 70
Body test, psi: 15,000

Class, psi: 15,000
MPa: 103.4
GOST series, MPa: 105
Body test, psi: 22,500

Class, psi: 20,000
MPa: 137.9
GOST series, MPa: 140
Body test, psi: 30,000

Class, psi	MPa	GOST series, MPa	Body test, psi
2,000	13.8	14	4,000
3,000	20.7	21	6,000
5,000	34.5	35	7,500
10,000	68.9	70	15,000
15,000	103.4	105	22,500
20,000	137.9	140	30,000

The classes and the factory test pressures of the body, with water, follow the standard API Spec 6A (ISO 10423) for flanges with a bore up to 13⅝″ (346 mm). For flanges of 16¾″ (425 mm) and larger at 2000 and 3000 psi, the test pressure is one and a half times the working pressure (3000 and 4500 psi). The metric series is GOST 30196 “Casing heads” (Kazakhstan was among the countries that adopted it). Its test pressure is also twice or one and a half times the working pressure, depending on the class and the bore. 1 MPa = 145.04 psi.

The factory test checks the strength of the body. On the rig, after each unit is installed, the crew tests the connections and the seals. The pressure must not be above the class of the unit, or above what the string can take. The pressure in the cavity between the seals presses on the pipe from the outside, and the collapse pressure of the pipe (rated characteristics) limits the test. The crew pressure-tests the preventers on a new spool with a test plug seated in its bowl (checks and pressure tests of BOP equipment).

Worked examplePressure classes of the wellhead of the lesson's well

Given (assumed data)

Surface casing: 339.7 × 12.19 mm to 700 m; leak-off pressure (the pressure at which the rock starts to take fluid) under the shoe, from the test, 1.50 g/cm³
Interval for the 244.5 mm string: 311.1 mm hole to 2605 m; normal pore pressure (the pressure of the fluid in the rock), 1.05 g/cm³
Interval below the 244.5 mm string: 215.9 mm hole to 4000 m; highest expected wellhead pressure 41.93 MPa (the string design example)
Gas; margin on the fracture pressure: 0.20 g/cm³; +0.06 g/cm³, as in the burst design of the string

Solution

Drilling for the 244.5 mm string, with the preventers on the lower casing head. Formation pressure at 2605 m: $0.00981\cdot 1.05\cdot 2605 = 26.83$ MPa. Under a column of gas down to the shoe of the surface casing, it would be $26.83 - 0.00981\cdot 0.20\cdot 1905 = 26.83 - 3.74 = 23.09$ MPa.
The rock under the shoe will hold no more than $0.00981\cdot(1.50 + 0.06)\cdot 700 = 10.71$ MPa. This is lower than the pressure that the gas column would give at the shoe, so fracture limits the pressure at the shoe. At the wellhead the pressure is $10.71 - 0.00981\cdot 0.20\cdot 700 = 10.71 - 1.37 = 9.34$ MPa. The class of 2000 psi (13.8 MPa) would be enough.
Drilling below the 244.5 mm string, with the preventers on the spool: 41.93 MPa is more than 5000 psi (34.5 MPa), so the class of 10,000 psi, 70 MPa, is needed.
The upper flange of the spool is 70 MPa. A spool with a seal on the string usually raises the class by only one step. So its lower flange must be at least 35 MPa, and the upper flange of the lower casing head must be the same, although 14 MPa would be enough for drilling for the 244.5 mm string.
Factory test of the body: for a head of 5000 psi, 7500 psi, that is 51.7 MPa; for a spool of 10,000 psi, 15,000 psi, 103.4 MPa (its lower part, under the secondary seal, is tested at the pressure for the lower flange).

Answer: The lower casing head is for 35 MPa (in GOST 30196, GK1K-340-350×35-245). Above it is a crossover spool of 35/70 MPa (GK2K-350×35-280×70). The preventers for drilling below the 244.5 mm string are for 70 MPa.

What it means. The class of the lower head was set not by its own job but by what will stand above it: one spool usually does not go from 14 straight to 70 MPa. So we choose the wellhead for the whole well at once, starting from the last string, in the same way as we choose the casing diameters. If the pressure while drilling for the 244.5 mm string were above 35 MPa, the lower head would have to be 70 MPa. The factory test checks only the body. On the rig the crew tests the connections and the seals separately, and at a lower pressure (an example is in the section on landing the string). The data in this example are assumed for illustration.

Preparing to run, and the rig equipment

The crew prepares the string long before running it. The pipes, the casing accessories (shoe, float collar, centralisers and the like) and the tools are delivered under a programme. For each section, the programme gives the wall thickness, the grade and the connection type (string sections). A pipe of another thickness or grade in the wrong place is a weak link that is hard to find later.

Before running

1
Receiving and laying out
The crew inspects the pipes, checks the markings and lays the pipes on racks in running order, section by section.
2
Drifting and the tally
The crew drifts each pipe (the drift), measures its length without the thread that goes into the coupling, numbers it and enters it in the tally of the string. From the tally the crew picks the pipes so that the shoe will stand at the right depth.
3
Threads
The crew takes the protector caps off the pins (the threaded ends of the pipes) and the couplings, cleans and inspects the threads, and sets aside pipes with damaged threads. The caps go back on, and come off only at the wellhead, just before the pipes are screwed together.
4
Hole and mud
The crew reams the hole (works the bit up and down to smooth and clean it) and checks it with a wiper trip (a trip of the drill string out and in to make sure the hole is clear). The crew then circulates the mud and treats it. The aim is to lower the yield point and the gel strength (two measures of how thick the mud is: while it flows and while it stands still), so that the string goes down freely, and the pressure while running and while breaking circulation (starting the pumps again) is lower (breaking the gel).
5
Casing accessories
The crew puts the shoe, the float collar, the centralisers and the stop rings (landing rings) on the lower pipes, as the cementing programme says (Primary cementing). The threads of the lower pipes are often locked with a special adhesive, so that drilling out these accessories does not unscrew them.
6
Equipment
The crew checks the elevators, the slips, the tongs with torque sensors, the fill-up tool, the crossover subs (short adapters between different threads) and the cementing head (the fitting on top of the string through which the cement is pumped). The crew pulls the wear bushing out of the casing head and inspects the bowl.

Equipment for running casing

Equipment: Single-joint elevator
Purpose: lifts a pipe from the catwalk to the wellhead

Equipment: String elevator, with a side door or a slip-type
Purpose: holds the string on the hoisting system; a slip-type elevator grips the pipe body and suits integral connections

Equipment: Spider: a slip-type grip on the rotary table
Purpose: holds the string while the next pipe is screwed on; the slips are long, so that they do not crush a thin wall

Equipment: Hydraulic power tong with a back-up tong
Purpose: screws the connection together; sensors record the torque and the turns

Equipment: Fill-up and circulating tool
Purpose: fills the string and lets the crew circulate the well without losing time to set up other equipment

Equipment: Casing running tool on the top drive (the motor on the derrick that turns the string)
Purpose: holds the string with the drive: the crew can rotate it, move it up and down and circulate while running

Equipment	Purpose
Single-joint elevator	lifts a pipe from the catwalk to the wellhead
String elevator, with a side door or a slip-type	holds the string on the hoisting system; a slip-type elevator grips the pipe body and suits integral connections
Spider: a slip-type grip on the rotary table	holds the string while the next pipe is screwed on; the slips are long, so that they do not crush a thin wall
Hydraulic power tong with a back-up tong	screws the connection together; sensors record the torque and the turns
Fill-up and circulating tool	fills the string and lets the crew circulate the well without losing time to set up other equipment
Casing running tool on the top drive (the motor on the derrick that turns the string)	holds the string with the drive: the crew can rotate it, move it up and down and circulate while running

The design and the names of particular systems differ among manufacturers. The lifting capacity of elevators and slips is chosen from the weight of the string, with a margin for overpull.

The crew lines the pipe up with the coupling exactly on axis, from a working platform at height or with a manipulator. A slant while starting the thread is one of the most common causes of a spoiled thread, and it is hard to notice until the connection begins to “bite” (to gall: the metal of the thread tears).

Screwing connections together: position, torque and the plot

A connection must hold the weight of the string and the pressure, and it must not let fluids through. Both depend on how it was screwed together (threaded connections). Before screwing the pipes together, the crew applies thread compound (thread grease) to the thread. It lowers friction and galling (tearing of the thread surfaces). In round and buttress threads it also fills the gaps along the spiral of the thread and makes the seal. Compounds are tested to the standard API RP 5A3 (ISO 13678). For premium connections the manufacturer sets the compound and its amount, and some connections come with a dry coating and no compound.

The crew first starts the connection at low speed, to make sure that the thread went in without a slant, and then tightens it with the power tong. What counts as correct tightening depends on the connection type:

Round thread (a triangular thread; API round: STC and LTC, with a short and a long coupling): Controlled by torque. The manufacturer or the standard gives the optimum torque with permitted limits. The position of the end of the coupling is checked as an additional sign.
Buttress thread (a trapezoidal thread; in GOST 632, OTTM): Controlled by position. The pin has a triangle mark. The end of the coupling must not go past the tip of the triangle, and it must not stop more than one turn short of the base of the triangle. The crew aims for the base. The torque at which this is reached on the first pipes then serves as a guide for the rest.
Premium connections with a torque shoulder: Controlled by the torque–turn plot. When the shoulders (the stop faces) meet, the torque rises sharply, and the final torque must fall in the window set by the manufacturer.

SchematicTorque–turn plot for screwing together a premium connection with a torque shoulder (a diagram). Correct tightening: a smooth rise of torque as the thread gets tight (interference), the shoulders landing, and a final torque inside the manufacturer's window. An early rise of torque means a slant or galling. A plot with no shoulder means that the connection is not tightened enough.

The plot shows what happened to the connection. A smooth rise of torque shows the thread getting tight (interference). For premium connections it is followed by a short stretch where the metal-to-metal sealing surfaces come together, and then by a sharp rise: the shoulders landing. A rise that is too early or uneven means that the thread went in at a slant or is galling. A final torque with no shoulder means that the connection is not tightened enough. A torque above the limit risks plastic deformation (a permanent change of shape). The crew unscrews such connections and inspects them, and replaces the pipe if needed.

Fill-up, buoyancy and hook load

The crew runs the string with a float valve (a one-way valve) in the shoe or in a collar just above it (parts of a string). After the displacement, the valve keeps the heavy cement slurry from returning into the string. But while the string is run, the valve also keeps mud out of the string. A string closed at the bottom pushes the whole of its outside volume out of the well, like a float. So the crew fills the string: every few pipes, the crew pumps mud into it from the top. Another way is an auto-fill float valve. While running, it is open, and mud enters the string from below with no overflow at the surface. Russian equipment calls this valve a “throttling check valve”. Before cementing, the crew switches it to the working position, by dropping a ball or by raising the pump rate above a set value.

Not filling the string is dangerous in three ways:

Buoyancy. A string that is empty inside weighs much less in mud. The mud pushes up on it with a force equal to the weight of the mud displaced by its whole outside volume. The weight indicator shows less than expected, and this is the first sign that the string is not full.
Collapse. The whole column of mud presses from outside on an empty string, and at a great depth the pressure can exceed the collapse pressure of the pipe (emptying of the string).
Loss of level. If the float valve of an unfilled string fails (cuttings jam it, or it is damaged), mud from the annulus rushes into the string. The level in the annulus drops, and the pressure on all the formations of the open hole falls. This is the same U-tube effect as when a heavy slug (a batch of heavy mud) is pumped (communicating vessels): the string and the annulus act as the two arms of a U-shaped tube. But here the pressure decreases, and a formation may start to flow.

Weight of a metre of string when the fluids inside and outside differ

w = g\left[m\left(1 - \frac{\rho_{\text{out}}}{\rho_{\text{steel}}}\right) + (\rho_{\text{in}} - \rho_{\text{out}})\,A_{\text{in}}\right]

where:

$w$ effective weight of a metre of string: the contribution of one metre to what the weight indicator shows, N/m
$m$ nominal weight (mass per metre) of the pipe with couplings, kg/m
$\rho_{\text{out}},\ \rho_{\text{in}}$ density of the fluid outside and inside (an empty string has $\rho_{\text{in}} = 0$ ), kg/m³
$\rho_{\text{steel}}$ density of steel, 7850 kg/m³, kg/m³
$A_{\text{in}}$ area of the inside cross-section of the pipe, m²
$g$ acceleration due to gravity, 9.81, m/s²

Units: SI

Holds when

A vertical string and fluids at rest. This is the same formula as in the lesson on drill string design (the effective force), where the volume of steel of the pipe with couplings is taken as $m/\rho_{\text{steel}}$ .
With the same fluids inside and outside, the second term disappears, and we get $w = m\,g\,k_{\text{b}}$ with the buoyancy factor $k_{\text{b}} = 1 - \rho_m/\rho_{\text{steel}}$ .
In an inclined hole the hook load also depends on the friction against the wall.

The first term is the weight of the pipe in the outside fluid, and the second is the weight of the fluid inside over and above the fluid outside. An unfilled 244.5 mm string in mud of 1.09 g/cm³ loses about 0.41 kN on every metre, almost three quarters of its weight in mud.

Drop of the level in the annulus when the float valve fails

y = L_{\text{empty}}\,\frac{A_{\text{in}}}{A_{\text{in}} + A_{\text{ann}}}, \qquad \Delta p = \rho_m\,g\,y

where:

$y$ drop of the level in the annulus; the level in the string rises to the same depth, m
$L_{\text{empty}}$ length of the unfilled upper part of the string, m
$A_{\text{in}},\ A_{\text{ann}}$ area of the inside cross-section of the string and area of the annulus at the wellhead, m²
$\Delta p$ fall of the pressure on the formations of the open hole, Pa
$\rho_m$ mud density, kg/m³

Units: SI; in metric units $\Delta p\,[\text{MPa}] = 0.00981\,\rho_m\,[\text{g/cm}^3]\,y\,[\text{m}]$

Holds when

The mud in the string and in the annulus has the same density, and the well is not being topped up at this time.
The levels equalise completely. The areas are constant along the unfilled part and along the depth of the drop.

The mud that leaves the annulus fills the string: $A_{\text{ann}}\,y = A_{\text{in}}(L_{\text{empty}} - y)$ . The larger the inside cross-section of the string is compared with the annulus, the larger the share of the unfilled length by which the level in the annulus falls.

Worked exampleHow often to fill the 244.5 mm string

Given (assumed data)

String: 244.5 mm, by the sections of the lesson on string design; pipes of 12 m; in the middle section the wall is 11.05 mm and the inside diameter is 222.40 mm
Annulus at the wellhead: inside the surface casing of 339.7 × 12.19 mm (inside diameter 315.32 mm)
Mud; pore pressure at 2600 m: 1.09 g/cm³; 1.05 g/cm³, still normal: here is the top of a zone of abnormally high formation pressure (overpressure)
Condition of the programme: a failure of the float valve must not take away more than half of the overbalance (the excess of the mud pressure over the formation pressure)

Solution

Areas: inside the string $\frac{\pi}{4}\cdot 0.2224^2 = 0.03885$ m², annulus $\frac{\pi}{4}(0.31532^2 - 0.2445^2) = 0.03114$ m². The share of the unfilled length by which the level falls: $0.03885/(0.03885 + 0.03114) = 0.5551$ .
Overbalance at 2600 m: $0.00981\cdot(1.09 - 1.05)\cdot 2600 = 1.020$ MPa, and half of it is 0.510 MPa. The allowed drop of the level is $y = 0.510/(0.00981\cdot 1.09) = 47.7$ m. The unfilled length must be no more than $47.7/0.5551 = 85.9$ m.
The crew must fill the string at least every 7 pipes (84 m). If the float valve fails, the level drops by $84\cdot 0.5551 = 46.6$ m, and the pressure by $0.00981\cdot 1.09\cdot 46.6 = 0.498$ MPa.
Hook load: each unfilled pipe holds $0.03885\cdot 12 = 0.466$ m³ of mud, so the string weighs less by $0.466\cdot 1090\cdot 9.81 = 4983$ N. Before a fill-up it is short of 34.9 kN.
If the crew did not fill the string at all, with the shoe at the bottom it would weigh 0.382 MN on the hook instead of 1.456 MN (26%). The mud column, $0.00981\cdot 1.09\cdot 2600 = 27.8$ MPa, would press from outside on its empty lower end, where the collapse pressure of the lower pipes is 32.8 MPa.

Answer: The crew fills the string at least every 7 pipes. Each unfilled pipe lowers the hook load by about 5.0 kN.

What it means. Collapse does not threaten this string, because it was designed for full emptying (collapse). So the fill-up here is needed first of all for well control: an unfilled string with a failed float valve would take the overbalance away from formations of normal pressure at the top of the overpressure zone. The hook load is the main gauge for the fill-up: every 12 m of empty string is about 5 kN. If the weight suddenly grows with no fill-up, the float valve has let mud in from the annulus. The crew fills the annulus at once and watches the well (trips). The data in this example are assumed for illustration.

CalculatedThe calculated hook load of the string of the example (without the travelling block, the moving block of the hoisting system) against the depth of the shoe while running: the string is filled all the time; filled every 7 pipes (the sawtooth: the weight drops while the pipes are empty and recovers at the fill-up); not filled at all.

The crew counts volumes while running with the trip tank, a small tank with a level gauge. Each 12 m pipe with a closed bottom pushes its outside volume out of the well, about 0.57 m³. When the crew fills it, 0.47 m³ goes into it. With an auto-fill float valve, only the volume of the steel is displaced, about 0.1 m³ per pipe. If the real volumes differ from the calculated ones, this is a sign of lost circulation (mud escaping into the rock) or of an inflow.

Running speed and pressure surges

As the string goes down, it pushes mud out, and the mud has to go up the annulus. Friction in the annulus raises the pressure on the walls of the open hole. We state this extra pressure as an equivalent mud weight (EMW): the density of mud that would give the same pressure at that depth. The lesson on pressure losses explains how to calculate it (running and pulling pipe). The same lesson gives the estimate that we use here. The speed of the mud in the annulus has two parts: the volume of the pipe that pushes the mud aside, and the mud that the pipe drags along (the coefficient $K_c$ ). We calculate the losses with the laminar formulas for a slot (flow in layers, with no turbulence). For casing the effect is stronger than for drill pipe, for two reasons. With a closed bottom, the casing pushes out its whole outside volume. And the gap between it and the wall is narrow.

Worked examplePressure surges while running the 244.5 mm string

Given (assumed data)

String and hole: 244.5 mm with a float valve, the shoe near the bottom (2600 m). The annulus to 700 m is inside the surface casing (315.32 mm). Below that, the hole by the caliper log (a log that measures the hole diameter) is 320 mm down to 2100 m and 316 mm deeper
Mud: 1.09 g/cm³; at 100 and 3 rpm the viscometer reads 12 and 3 divisions: $n = 0.395$ , $K = 0.804$ Pa·sⁿ (the power-law model)
Running speed: on average about 20 pipes an hour (0.067 m/s), but the crew lowers each pipe in a few tens of seconds, so the highest speed is about 0.5 m/s; $K_c = 0.45$
Margin to the leak-off pressure: under the surface casing shoe $1.50 - 1.09 = 0.41$ g/cm³; at the bottom $1.784 - 1.09 = 0.69$ g/cm³

Solution

Mud speed in the 320 mm hole: $\dfrac{244.5^2}{320^2 - 244.5^2} = 1.4026$ , $v = 0.5\cdot(1.4026 + 0.45) = 0.9263$ m/s. Inside the surface casing it is 0.9789 m/s, and in the 316 mm hole 0.9709 m/s.
Losses in the 320 mm hole: the shear rate at the wall is $\dfrac{2n+1}{3n}\cdot\dfrac{12v}{D-d} = 1.5105\cdot\dfrac{12\cdot 0.9263}{0.0755} = 222.4$ s⁻¹. The stress is $\tau_w = 0.804\cdot 222.4^{0.395} = 6.80$ Pa. The gradient is $4\cdot 6.80/0.0755 = 360.3$ Pa/m. In the same way, inside the surface casing it is 402.8 Pa/m, and in the 316 mm hole 396.1 Pa/m. The flow is laminar everywhere.
Added pressure at the shoe: $402.8\cdot 700 + 360.3\cdot 1400 + 396.1\cdot 500 = 0.984$ MPa, which is an equivalent mud weight of $0.984/(0.00981\cdot 2600) = 0.039$ g/cm³. Under the surface casing shoe there are only the losses above it: $402.8\cdot 700 = 0.282$ MPa, 0.041 g/cm³.
Drill pipe of 127 mm at the same speed: the area ratio is 0.1870 instead of 1.4026, the gap is 2.6 times wider, and the added pressure at the bottom is 0.169 MPa (0.0066 g/cm³). This is $0.984/0.169 = 5.8$ times less.
Breaking circulation after the run: mud with a gel strength of 4 Pa after 10 min at rest needs $4\cdot 4\cdot\left(\dfrac{700}{0.0708} + \dfrac{1400}{0.0755} + \dfrac{500}{0.0715}\right) = 0.57$ MPa, which is another 0.022 g/cm³ at the bottom.

Answer: At 0.5 m/s the pressure at the bottom grows by 0.984 MPa (0.039 g/cm³), and under the surface casing shoe by 0.041 g/cm³. The margins to the leak-off pressure are 10–18 times larger.

What it means. In this well the window is wide. What limits the running speed is not the pressure surges but the time for screwing the pipes together and filling them, ledges in the hole, and the shock loads of a sudden stop (additional loads). While the losses in the gap follow the laminar formula, the dependence on speed is weak. A mud with $n \approx 0.4$ behaves almost like a plastic body (one that flows only when the stress is high enough), and at 0.8 m/s the added pressure is only 20% larger than at 0.5 m/s. The Reynolds number in the gap passes the laminar limit already at about 0.7 m/s. But the turbulent losses overtake the laminar ones only above about 0.9 m/s. Beyond that the added pressure grows more steeply: at 1.5 m/s it is already 2.6 times larger than at 0.5 m/s (the figure below). On the other hand, a string with a closed bottom loads the walls several times harder than drill pipe: it pushes its whole outside volume through a gap that is 2.6 times narrower. In a narrow gap everything is different. A 177.8 mm liner in a 215.9 mm hole at the same speed gives a gradient of about 1063 Pa/m, almost three times more than in the 320 mm hole. If the window between the pore pressure and the leak-off pressure is narrow, the running speed must be calculated. The data in this example are assumed for illustration.

CalculatedThe calculated added pressure at the bottom (as an equivalent mud weight, EMW) against the running speed: a 244.5 mm string with a closed bottom; the same string with an auto-fill float valve, which is the lower estimate, when the whole displaced volume goes inside the string; 127 mm drill pipe. The margins to the leak-off pressure in the lesson's well (0.41 and 0.69 g/cm³) lie far above the top edge of the graph.

Landing the string, cutting off and the next spool

When the string has reached its planned depth, the crew circulates and cements it (Primary cementing), and lands it in the casing head in one of three ways:

Land, then cement. A mandrel hanger sits in the bowl before cementing, and the string hangs in it during the job.
Cement while hanging, then set the slips. The string hangs on the elevator. After the displacement, the crew sets the slips with the weight that the weight indicator shows. Or, if needed, after the cement sets, the crew pulls on the string and sets it with extra tension. The extra tension keeps the heating while drilling and producing from compressing its uncemented part (after cementing).
Land with some weight set down (slack-off). The crew lowers the string a little and leaves part of its weight on the cement. This is rarely done, when tension reduces the collapse resistance too much (biaxial loading).

The hook load changes during cementing. While heavy slurry is in the string, the string gets heavier. When the slurry goes out into the annulus, a fluid heavier than mud pushes up on the string from the outside, and it becomes lighter than it was before cementing. Adding up the corrections of the weight formula along the whole length, we get:

Hook load of a string with fluids of different densities

W = W_m - g\,A_{\text{out}}\sum_j (\rho_{\text{out }j} - \rho_m)\,h_j + g\,A_{\text{in}}\sum_k (\rho_{\text{in }k} - \rho_m)\,h_k

where:

$W$ hook load of the string (without the travelling block and the drive), N
$W_m$ weight of the string in mud, with mud inside and outside, N
$A_{\text{out}}$ outside area that displaces fluid: $A_{\text{in}} + m/\rho_{\text{steel}}$ , allowing for the steel of the couplings, m²
$\rho_{\text{out }j},\ h_j$ density and height of the $j$ -th column of fluid behind the string, kg/m³, m
$\rho_{\text{in }k},\ h_k$ density and height of the $k$ -th column inside the string, kg/m³, m
$\rho_m$ density of the drilling mud, kg/m³

Units: SI; it is convenient to calculate $\sum(\rho - \rho_m)h$ in g/cm³·m and to multiply by 1000

Holds when

A vertical string and fluids at rest. When mud circulates, the friction of the flow against the pipe adds to the weight.
The areas are the same along the length. For a string of sections with different thickness, $A_{\text{out}}$ hardly changes, and $A_{\text{in}}$ is taken for the section where the column stands.

Each metre of a column of fluid heavier than the mud behind the string reduces the weight by $g A_{\text{out}}(\rho - \rho_m)$ , and inside the string it increases the weight by $g A_{\text{in}}(\rho - \rho_m)$ .

Worked exampleAt what weight to land the string and at what pressure to test the seals

Given (assumed data)

String in mud before cementing: hook load 1456.2 kN (tension design of the string)
Behind the string after the displacement: mud down to 50.4 m, wash fluid (a fluid that cleans the hole ahead of the cement) of 1.00 g/cm³ for 128.5 m, spacer (a fluid that separates mud from cement) of 1.25 g/cm³ for 321.1 m, lead slurry (the first batch of cement) of 1.45 g/cm³ for 1600 m, tail slurry (the last, heavier batch) of 1.90 g/cm³ for 500 m; inside is mud, and in the shoe track (24 m) is tail slurry
Areas: $A_{\text{in}}$ = 0.0382 m² (wall 11.99 mm); $A_{\text{out}} = A_{\text{in}} + m/\rho_{\text{steel}}$ ≈ 0.0471 m²
Top pipe: 244.5 × 11.99 mm, N-80: body tensile strength 4.83 MN, collapse pressure 32.8 MPa
Test through the test port: the flange under the spool is 35 MPa; the programme limits the pressure on the pipe from outside to 0.8 of its collapse pressure

Solution

Fluids heavier than the mud push up on the string from outside: $\sum(\rho_{\text{out}} - 1.09)\,h = -0.09\cdot 128.5 + 0.16\cdot 321.1 + 0.36\cdot 1600 + 0.81\cdot 500 = 1020.8$ g/cm³·m.
This reduces the weight by $9.81\cdot 1000\cdot 0.0471\cdot 1020.8 = 471.7$ kN. The tail slurry in the shoe track adds $9.81\cdot 1000\cdot 0.0382\cdot 0.81\cdot 24 = 7.3$ kN.
Hook load at the end of the displacement: $1456.2 - 471.7 + 7.3 = 991.8$ kN. The crew lands the string on the slips with this weight. The hanger will carry about 0.99 MN, and at the top pipe the margin to yield is $4.83/0.99 = 4.9$ .
Test of the flanged joint and the seals: not above 35 MPa and not above $0.8\cdot 32.8 = 26.2$ MPa. The crew takes 26 MPa.

Answer: The crew lands the string on the slips with a weight of about 0.99 MN, which is 32% less than its weight in mud. It tests the seals and the flanged joint at 26 MPa.

What it means. When the cement sets, the weight with which the string was landed is “frozen” into its uncemented part. Suppose the calculation shows that heating while drilling deeper, or while producing, will take away this tension. Then, while the cement is liquid, the crew holds the string on the elevator, and after the cement sets, the crew pulls the string and lands it with extra tension. A test at the pressure of the flange class (35 MPa) would exceed the collapse pressure of the top pipe, so the test was limited. The data in this example are assumed for illustration.

Landing on the slips and installing the spool

1
Landing
After the displacement and a check that the float valve holds, the crew disconnects the preventers from the casing head and lifts them. The crew puts the slip hanger around the string in the bowl, and smoothly lowers the weight of the string onto it, up to the weight set by the programme.
2
Rough cut
The crew cuts the string above the head and removes the cut-off pipe. The crew takes the preventers off.
3
Final cut and chamfer
The crew cuts the string square, at the height above the flange that the spool manufacturer sets. The cut-off end must enter the secondary seal. The crew chamfers (bevels) the outside edge and removes burrs (sharp metal edges) on the inside, so that the seal goes on without damage.
4
Seal and spool
The crew installs the hanger seal, a new ring gasket and the spool. The secondary seal sits on the end of the string. The studs are tightened in the set order.
5
Tests
Through the test port the crew pressure-tests the flanged joint and the seals (the example above). The crew puts the preventers on the upper flange of the spool and pressure-tests them with a plug in the bowl (pressure tests of BOP equipment). The crew installs the wear bushing.
6
Next
After waiting on cement (WOC, the pause while the cement hardens), the crew drills out the casing accessories and the shoe. The crew then drills a few metres of new rock and does a test below the shoe (LOT and FIT, the leak-off test and the formation integrity test) to find how much pressure the rock can take.

A liner hangs not in the casing head but inside the previous string. The crew runs it on drill pipe and sets it on a liner hanger with slips that are pushed out mechanically or by pressure (liner). The wellhead does not change. The lesson on primary cementing explains how a liner is cemented.

Terms and symbols

Casing wellhead: A stack of flanged units (casing heads, spools, the tubing head) in which the strings hang and the annuli between strings are closed.
Casing head: The lower one is single-flanged, on the surface casing. The following ones are double-flanged spools (casing spools), one for each string.
Crossover spool: A spool with a secondary seal on the string, whose upper flange is usually one step higher in class than the lower one.
Casing hanger: A mandrel (threaded) hanger is screwed onto the top pipe. A slip hanger has slips that grip the pipe at any point.
Primary and secondary seals: The hanger seal under the spool, and the spool seal on the cut-off end of the string. Between them is a cavity with a test port.
Wear bushing: A bushing that covers the bowl of the head while the crew drills through it.
Tally and spacing out: The list of pipes with their measured lengths, and the selection of them (with pup joints) so that the shoe stands at the right depth.
Filling the string: Filling a string that has a float valve with mud from the top, every few pipes.
Auto-fill float valve: A float valve that is held open while the string runs, so that mud enters the string from below. The crew closes it before cementing.
Torque–turn plot: A record of screwing a connection together that shows the thread getting tight, the meeting of the shoulders and the final torque.
Surge pressure: The addition to the hydrostatic pressure from the friction of the mud that the string pushes out into the annulus.

Common mistakes and misconceptions

Wrong

“A lower casing head of 35 MPa, when 10 MPa is expected, is an unnecessary margin.”

In fact

The class also sets what will stand above the head. A crossover spool usually raises the class by only one step, so for a spool of 70 MPa, the flange below it must be 35 MPa.

Wrong

“The float valve holds, so the fill-up can wait until the end of the run.”

In fact

An empty string floats, and its bottom is loaded by the external pressure. If the float valve fails, the level in the annulus drops and the formations lose their overbalance. The crew fills the string by calculation: in the example of the lesson, at least every 7 pipes.

Wrong

“The tighter the connection, the more reliable.”

In fact

A torque above the limit deforms the thread and the seals. For a buttress thread, the main sign is the position of the coupling relative to the triangle. For a premium thread, it is the shape of the torque–turn plot.

Wrong

“A string can be run as fast as drill pipe.”

In fact

A string with a closed bottom pushes out its whole outside volume through a narrow gap. At the same speed, the pressure on the walls grows several times more.

Wrong

“After cementing the string weighs the same as before it.”

In fact

The heavy slurry behind the string pushes up on it. In the example of the lesson, the hook load fell by a third. The crew lands the string on the slips with this weight, unless extra tension is needed.

In depth: advanced detail

A string with an auto-fill float valve: between two estimates

In a string with an open bottom, the mud displaced by the steel splits into two streams: part goes up the annulus, and part goes inside the string. It splits so that the pressure drop from the bottom of the string to the surface is the same by both paths. The inside channel is wide (222 mm), and the annulus is a slot 35–38 mm wide. So almost all of the mud goes inside, and the pressure surges are much smaller than with a closed bottom.

Simple estimates give the bounds. The upper bound is a string with a closed bottom (the first term of the speed formula). The lower bound is the case when the whole displaced volume has gone inside, and only the mud that the string drags along stays in the annulus: $v_{\text{ann}} = K_c\,v_{\text{pipe}}$ . For the lesson's string at 0.5 m/s these are 0.039 and 0.022 g/cm³ (as equivalent mud weight) at the bottom. Even with no displacement of mud through the annulus, the surges stay noticeable, because the wall of the slot itself is moving. The true value lies between the bounds, and closer to the lower one. The resistance of the valve itself at high speed raises it.

Connections to other topics

Casing: purpose, types, properties and standards — strings by purpose, the parts of a string, the drift, threaded connections and the rated characteristics of pipes.
Casing design — the string sections of the lesson, the wellhead pressure of a closed well, tension and the loads after cementing.
Blowout prevention equipment — the preventers installed on casing heads, their working pressure and pressure tests.
Drill string design — the weight of a string in mud, the effective force and the hook load.
Circulating pressure losses — where the pressure surges while running come from, and how to calculate them.
Well control: primary barrier and kick indicators — communicating vessels, fill-up and volume control during trips.
Primary cementing — how the string is cemented after running: the same well and the same string.
Completion string equipment — the tubing head, the tubing hanger and the Christmas tree: the continuation of the casing wellhead.

Check yourself

1
The lower flange of a crossover spool is rated for 35 MPa. What class can its upper flange usually have?
1. A.105 MPa
2. B.Only 35 MPa
3. C.Any: the class is set by the studs
4. D.70 MPa
Show the answer
Correct answer: D. 70 MPa
The secondary seal on the cut-off end of the string separates the well pressure from the lower flanged joint, and the pressure acts on that joint over a small area only. This allows a higher class for the upper flange, usually by one step: 35 → 70 MPa. So for a spool of 70 MPa, the lower casing head must be at least 35 MPa.
2
A 244.5 mm string is run with a float valve, the 10 upper pipes of 12 m are not filled, and the valve has failed. By how much will the pressure on the formations fall (mud 1.09 g/cm³, $A_{\text{in}}$ = 0.0388 m², $A_{\text{ann}}$ = 0.0311 m²), and how much of the overbalance of 1.02 MPa is left?
Show the answer
The unfilled length is 120 m. The level in the annulus falls by $y = 120\cdot 0.0388/(0.0388 + 0.0311) = 120\cdot 0.5551 \approx 66.6$ m, and the pressure by $0.00981\cdot 1.09\cdot 66.6 \approx 0.71$ MPa. About 0.31 MPa of the overbalance is left. A few more unfilled pipes, and formations of normal pressure may start to flow. So the crew calculates the fill-up interval, and fills the annulus at once after such a failure.
3
Why, at the same running speed, are the pressure surges of a casing string with a float valve larger than those of drill pipe?
1. A.Mud is always thicker when a string is run
2. B.It pushes out its whole outside volume, and the gap between it and the hole is several times narrower
3. C.The string is heavier
4. D.Casing has a rougher surface
Show the answer
Correct answer: B. It pushes out its whole outside volume, and the gap between it and the hole is several times narrower
The speed of the mud in the annulus is proportional to $d^2/(D^2 - d^2) + K_c$ . For a 244.5 mm string in a 320 mm hole, the ratio of the areas is about 1.40, and for 127 mm drill pipe it is 0.19. In addition, the gap of the string is 2.6 times narrower, and losses in a slot grow as its width falls. In the example of the lesson, the added pressure from the string is almost 6 times larger.
4
A 244.5 mm string is run with a float valve. The last 5 pipes (60 m) are not yet filled, and suddenly the weight indicator shows noticeably more than expected. What has probably happened, how far has the level in the annulus fallen, and what should the crew do ( $A_{\text{in}}$ = 0.0388 m², $A_{\text{ann}}$ = 0.0311 m², mud 1.09 g/cm³)?
Show the answer
Probably the float valve has failed. Mud has flowed from the annulus into the string, the levels inside and outside have equalised, and the string has become heavier: by $1090\cdot 9.81\cdot 0.0388\cdot 60 \approx 25$ kN, which it was short of compared with a full string, and by a little more, because the top of the pipe is now above the mud, and the mud no longer pushes up on it there. The level in the annulus has fallen by $60\cdot 0.0388/(0.0388 + 0.0311) \approx 33$ m, and the pressure on the formations by $0.00981\cdot 1.09\cdot 33 \approx 0.35$ MPa, which is about a third of the overbalance of 1.02 MPa in the example of the lesson. The crew fills the annulus at once, checks for flow, and runs on knowing that the valve does not hold: after the displacement, it will have to hold the pressure in the string until the cement sets.
5
How is the screwing together of a buttress thread on casing controlled?
1. A.By the pressure test of the string
2. B.Only by the torque on the power tong
3. C.By the position of the end of the coupling relative to the triangle mark on the pin; the torque serves as a guide
4. D.By the number of turns made by hand
Show the answer
Correct answer: C. By the position of the end of the coupling relative to the triangle mark on the pin; the torque serves as a guide
For a buttress thread, the main sign is position: the end of the coupling must not go past the tip of the triangle, and it must not stop more than one turn short of its base. The torque at which this is reached on the first pipes is then used as a guide. A round thread is controlled by torque, and premium connections by the torque–turn plot.
6
Why were the seals and the flanged joint under the spool in the example of the lesson tested at 26 MPa and not at 35 MPa, the flange class?
Show the answer
The test pressure acts in the cavity between the seals on the pipe from the outside. The collapse pressure of the top pipe, 244.5 × 11.99 mm of grade N-80, is about 32.8 MPa, and 35 MPa would exceed it. The programme limits the test to a share of the collapse pressure (0.8 in the example), and the smaller of the two values sets the limit: the flange class or what the pipe can take.

Updated: 28 October 2026