Remedial cementing, cement plugs and cement evaluation

Quick refresher

We check the cement behind the casing only indirectly. A pressure test checks the casing itself. A test below the shoe (the bottom end of the casing) checks the cement at the shoe. A temperature survey in the first day finds the top of the cement. Acoustic and ultrasonic tools show how well the cement is bonded and whether it fills the annulus (the space behind the casing). The final answer on sealing comes from a pressure test applied in the direction in which the formation will push.
Free casing (casing with no bonded cement) “rings”: the wave along it is strong. Cement bonded to the pipe weakens the vibration. The bond index is a ratio of attenuations, not of amplitudes. An amplitude 6 times lower than in free casing can mean that only half of the circumference is bonded.
A low amplitude does not yet mean a seal: the tool may have moved against the wall, or the cement may not be bonded to the rock. A high amplitude does not yet mean a channel: it may be a microannulus, fast rock or lightweight cement. So we look at the travel time and the wave pattern, and if in doubt we repeat the log under pressure.
Cement does not enter the pores of the rock. In a squeeze without fracturing, the slurry gives up its filtrate (the water that passes into the rock) and leaves a dense cake in the perforations. We keep the pressure on the formation below the fracture pressure. The allowed wellhead pressure falls while the heavy slurry fills the pipe.
A balanced plug has the cement at the same height inside the pipe and in the annulus. The spacer behind the cement in the pipe has the same height as the spacer ahead of it in the annulus. We pump slightly less displacement than the full volume. Pumping extra pushes cement out of the pipe into the annulus and upsets the balance. A bigger excess pushes spacer and mud under the plug.
Under the rules of Kazakhstan, the plug in an ordinary well covers the perforated interval, or a formation that holds oil, gas or water, with a margin of 20 m above and below. In an open hole with no production casing, the plug reaches at least 50 m above the top of the highest formation. We test plugs by setting weight on them. A plug at the shoe of the last intermediate casing (a string run to isolate problem intervals) is also pressure-tested.

Key relationships

\text{BI} = \dfrac{\log_{10}(A_{\text{free}}/A)}{\log_{10}(A_{\text{free}}/A_{100})}

bond index from the amplitude of the casing wave: $A_{\text{free}}$ is the amplitude in free casing, $A_{100}$ the amplitude in an interval of full bond with the same cement

h_{\text{cem}} = \dfrac{V_{\text{cem}}}{A_{\text{ann}} + A_{\text{pipe}}}, \qquad V_{\text{sp2}} = V_{\text{sp1}}\,\dfrac{A_{\text{pipe}}}{A_{\text{ann}}}

balanced plug: the height of the cement while the pipe is in the hole, and the volume of the spacer behind the cement

p_{\text{wh}}^{\max} = p_f - \Delta p_{\text{margin}} - 0.00981\sum \rho_i h_i

limit of the wellhead pressure in a squeeze without fracturing, MPa ( $\rho$ in g/cm³, $h$ vertical, in m)

\Delta r = \dfrac{p\,r^2}{E\,t} + \alpha\,\Delta T\,r

growth of the casing radius with pressure and temperature: this is where a microannulus comes from

Common pitfalls

Calculating the bonded share from the ratio of amplitudes. An amplitude of 16% of that in free casing is not “84% bond”.
Taking a low amplitude with a shortened travel time as a good bond. The tool is lying against the wall.
Declaring an interval uncemented because of a high amplitude without checking for a microannulus with a log run under pressure. The same applies to fast rock (check the travel time and the wave pattern) and to lightweight cement (check against its own reference).
Testing a squeeze only by pressure from inside. Dehydrated cement in the perforations can hold pressure from inside and still fail under the pressure of the formation from outside.
Pumping extra displacement when a plug is set, and setting a heavy plug on a liquid mud with no support under it.

After this lesson you can

list the defects of the cement sheath and the methods that find them (a pressure test, a temperature survey, a cement bond log with its wave pattern, and ultrasonic tools), and say what each method can see and what it cannot see
calculate the attenuation (the loss of signal strength) and the bond index from the amplitude of a cement bond log, and recognise false readings: a microannulus (a hairline gap between the casing and the cement), an off-centre tool, fast rock (rock in which sound travels quickly), and cement that is not bonded to the rock
choose a squeeze method (without or with fracturing the rock; through the wellhead, with a packer, which is a rubber seal that blocks the annulus, or through a retainer, which is a drillable plug set in the casing; continuous or with pauses) and calculate the highest allowed wellhead pressure as the pipe fills with slurry
calculate a balanced cement plug (one that stands at the same height inside the pipe and in the annulus), including the volume of slurry, the spacer (a separating fluid) behind the cement, the displacement (the fluid that pushes the cement out of the pipe) and the height of the plug with the pipe in place and after it is pulled out, and explain why plugs fail and how they are tested
design the plugs for abandoning a well under the rules of Kazakhstan now in force, and explain how abandonment differs from suspension (temporary closure of a well)

Defects of the cement sheath and how we find them

Primary cementing (see Primary cementing) should leave an unbroken ring of set cement behind the casing, from the shoe (the bottom end of the casing) up to the planned height. The ring must be bonded tightly both to the pipe and to the rock. This does not always happen. The defects differ in where they come from and in what they connect.

SchematicCross-sections of the well, looking along its axis: cement bonded to both the casing and the rock; a channel of drilling mud on the narrow side of an off-centre casing; a microannulus between the casing and the cement; cement bonded to the casing but separated from the rock by a mudcake (a thin layer of mud solids left on the wall of the hole). The gaps are exaggerated, and the sketch is not to scale.

Defects of the cement sheath

Defect: Top of cement too low
Where it comes from: lost circulation (mud escaping into the rock) during cementing, a hole wider than planned, an error in the volume
How it shows: an uncemented interval opposite formations that must be kept apart

Defect: Mud channel
Where it comes from: poor displacement of the mud on the narrow side of an off-centre casing (centralisation)
How it shows: flow between formations, water or gas at the perforations

Defect: Gas channel
Where it comes from: gas entered the cement while it was thickening (gas migration)
How it shows: pressure between the casing strings, gas seeping out at the surface near the well

Defect: Microannulus
Where it comes from: the casing shrank after the cement set, because the pressure or the temperature inside fell
How it shows: the acoustic amplitude reads too high although the cement is sound; sometimes a path for gas

Defect: No bond to the rock
Where it comes from: mudcake, or water that separated from the cement
How it shows: flow along the wall of the hole even though the bond to the pipe is good

Defect: Poor cement at the shoe
Where it comes from: contaminated slurry, or over-displacement (too much displacement fluid pumped)
How it shows: a low result in the test below the shoe

Defect	Where it comes from	How it shows
Top of cement too low	lost circulation (mud escaping into the rock) during cementing, a hole wider than planned, an error in the volume	an uncemented interval opposite formations that must be kept apart
Mud channel	poor displacement of the mud on the narrow side of an off-centre casing (centralisation)	flow between formations, water or gas at the perforations
Gas channel	gas entered the cement while it was thickening (gas migration)	pressure between the casing strings, gas seeping out at the surface near the well
Microannulus	the casing shrank after the cement set, because the pressure or the temperature inside fell	the acoustic amplitude reads too high although the cement is sound; sometimes a path for gas
No bond to the rock	mudcake, or water that separated from the cement	flow along the wall of the hole even though the bond to the pipe is good
Poor cement at the shoe	contaminated slurry, or over-displacement (too much displacement fluid pumped)	a low result in the test below the shoe

What a defect does depends on what it connects. A channel between two water-bearing formations is harmless. A channel from a water-bearing formation to an oil perforation makes the well produce water. A gas channel to the wellhead makes the well dangerous. So we read the results of the evaluation together with the geological section (the column of rock layers).

No method measures the seal of the sheath directly. Each one sees something different, so we combine them with the record of the job itself: the volumes, the pressures and the returns, which is the fluid coming back to the surface (monitoring during cementing).

Casing pressure test: done at the bump (when the top plug lands and the pump pressure rises) or later, before the shoe is drilled out. It checks the pipes, the threaded joints and the casing equipment, but not the cement behind the casing.
Test below the shoe: done after the shoe is drilled out (a leak-off test). If fluid leaks away at a pressure well below what is expected for these rocks, one possible cause is a channel in the cement at the shoe. The fluid goes along the channel behind the casing instead of splitting the rock.
Temperature survey: cement gives off heat as it hydrates, that is, reacts with water (clinker and hydration). A thermometer run into the casing in the first hours to the first day after cementing shows a step in temperature at the top of the cement. Opposite washouts (places where the hole is enlarged) there is more cement, and it is warmer there. A lightweight or heavily retarded (slow-setting) lead slurry, which is the first portion pumped and fills the top of the interval, warms up less and later. So we choose the time of the survey according to the slurry recipe.
Radioactive tracers: a radioactive tracer is added to the slurry, and a gamma ray log (a log of the natural radioactivity of the rocks) run before and after cementing is compared. The top of the cement is where the radiation falls back to the background level. The method is rarely used, because work with radioactive materials needs special permits and safety measures.
Gamma-gamma cement logging: used in the practice of the countries of the Commonwealth of Independent States (CIS). A tool with a source of gamma rays and detectors facing different ways measures the density of the material behind the casing in several directions, and the wall thickness of the pipe. Cement is heavier than mud, so we can see whether the annulus is filled and on which side mud is left. This method does not show the bond.
Acoustic and ultrasonic cement logging: the main methods. They show how well the cement is bonded to the pipe and to the rock and, with ultrasonic tools, how the bond is spread round the circumference. The next sections deal with them.

What the rules in Kazakhstan require

The Rules for ensuring industrial safety at hazardous production facilities of the oil and gas industries (order of the Minister for Investment and Development of the Republic of Kazakhstan No. 355 of 30 December 2014) require the following. After the casing is run, cemented and tested for strength and tightness, reports must be drawn up with the conclusions of the well logs on the condition of the casing and the cement sheath (paragraph 118). If the wellhead, the production casing or the intermediate casing strings leak, measures must be taken to remove the defects before the well is tested and brought into production. The same applies to flow between formations, pressure between the casing strings, and cemented intervals that do not match the requirements (paragraph 493). The pressure between the casing strings must be monitored through the whole life of the well (paragraph 469-1). Treatments of the near-wellbore zone and well stimulation (work to raise the flow) are not allowed in a well with a leaking wellhead or leaking casing, flow behind the casing or pressure between the strings (paragraph 736). The text of the rules is given as of October 2026.

Cement bond log: attenuation and bond index

The tool for a cement bond log (CBL) is built like the sonde (the logging tool) of a sonic log, which records how fast sound travels through rock (how the interval transit time is measured). A transmitter sends out a pulse with a frequency of about 20 kHz. Receivers stand about 0.9 m (3 ft) and 1.5 m (5 ft) from it. The wave goes through the mud to the casing, runs along the wall of the pipe and comes back to the receiver. Sound travels fast along the steel wall, about 187 µs/m (57 µs/ft) against 650 µs/m in mud, so the wave along the casing arrives first. The tool is held in the centre of the pipe (centred). If it lies against the wall, it distorts both the time and the amplitude (the height of the signal). The advanced section deals with this.

Free casing, with mud inside and outside, rings like a bell. Almost no energy leaves the pipe, and a strong signal reaches the receiver. If cement is bonded to the pipe, the vibration of the wall passes into the set cement and leaks away into it. The wave along the casing then gets weaker with every metre of its path. The stronger the cement and the larger the part of the circumference that is bonded to it, the stronger the attenuation. The cement weakens the wave less in a thick wall than in a thin one.

Three quantities are recorded. The travel time is the time to the first arrival at the near receiver. It tells us whether the tool is centred and whether there are fast rocks (rocks in which sound travels quickly). The amplitude is the height of the first positive half-cycle of the casing wave at the same receiver, in millivolts. The wave pattern of the far receiver is shown as a variable density log (VDL): positive half-cycles are dark and negative ones are light, with time along the horizontal axis and depth along the vertical axis.

Attenuation of the casing wave

\alpha = \frac{20}{\Delta s}\,\log_{10}\frac{A_1}{A_2}, \qquad \Delta\alpha = \frac{20}{L_{\text{eff}}}\,\log_{10}\frac{A_{\text{free}}}{A}

where:

$\alpha$ attenuation of the casing wave over the stretch between the two receivers, dB/m
$A_1,\ A_2$ amplitudes of the casing wave at the near and at the far receiver, mV
$\Delta s$ distance between the receivers, m
$\Delta\alpha$ attenuation that the cement added to the losses of the free pipe, found from one receiver, dB/m
$A_{\text{free}},\ A$ amplitude at the near receiver in free casing of the same well and in the interval under study, mV
$L_{\text{eff}}$ path of the wave along the casing to the near receiver: the 0.914 m spacing minus the shift caused by the two stretches of path through the mud, which gives about 0.9 m, m

Units: dB/m; amplitudes in the same units; 1 dB/ft = 3.28 dB/m

Holds when

The amplitude of the casing wave decays exponentially: on every metre of path it falls by the same factor.
Both receivers see the same casing wave, and the losses on the stretches of path through the mud are equal for them. So these losses cancel in the ratio $A_1/A_2$ .
In the second formula the reference is free casing in the same well: the same diameter and wall thickness of the pipe, the same mud and the same tool.

Decibels are handy because attenuations over successive stretches add up: $\Delta\alpha$ is what the cement added to the losses of the pipe itself. A measurement with two receivers does not depend on the strength of the transmitter pulse (and a layout with two transmitters does not depend on the sensitivity of the receivers either). With a single amplitude, we find the attenuation relative to a reference: free casing.

Bond index

\text{BI} = \frac{\Delta\alpha}{\Delta\alpha_{100}} = \frac{\log_{10}(A_{\text{free}}/A)}{\log_{10}(A_{\text{free}}/A_{100})}

where:

$\text{BI}$ bond index: 0 is free casing, 1 is full bond, —
$\Delta\alpha_{100}$ added attenuation in an interval of full bond of the same cement in the same casing, dB/m
$A_{100}$ amplitude in that interval: the lowest steady amplitude of this cement, mV

Units: dimensionless; amplitudes in mV

Holds when

The reference $A_{100}$ is taken in the same casing and in the same cement. A lightweight set cement weakens the wave less even with full bond, so the lead slurry needs its own reference.
The index averages over the whole circumference: a channel over part of the circumference and a uniformly weak cement give the same BI.
The travel time is constant, and arrivals through the rock do not fall into the window where the amplitude is measured (otherwise the readings are false; see the next section).

In a simple model of a channel, the cement weakens the wave equally on every bonded stretch, so the attenuation is proportional to the bonded share of the circumference, and we read BI as that share. Isolation (a seal between formations) is often considered likely if BI is at least about 0.8 over a continuous interval of several metres. This is a rule of experience. The criteria for the conclusions are set by the operator and the service company.

Worked exampleBond index in a production well

Given (assumed data)

Casing and cement: 168.3 × 8.94 mm. Opposite the productive part: class G cement (a standard oilwell cement) as the tail slurry (the last portion pumped). Above it: a lightweight lead slurry (the first portion pumped)
Free casing (above the top of the cement): $A_{\text{free}}$ = 61 mV
Best interval of the tail slurry: $A_{100}$ = 1.6 mV
Shale barrier at 2326–2342 m between the perforated intervals: 2.4 mV
Above the upper interval, 2296–2312 m: 9.5 mV
In the lead slurry: 24 mV; its best interval is 8 mV

Solution

Reference: $\log_{10}(61/1.6) = 1.5812$ . The added attenuation of full bond is $\Delta\alpha_{100} = (20/0.9)\cdot 1.5812 = 35.14$ dB/m, which is 10.71 dB/ft.
Barrier: $\log_{10}(61/2.4) = 1.4051$ , $\text{BI} = 1.4051/1.5812 = 0.89$ . The amplitude is 3.9% of that in free casing.
Above the upper interval: $\log_{10}(61/9.5) = 0.8076$ , $\text{BI} = 0.8076/1.5812 = 0.51$ , even though the amplitude is only 16% of that in free casing.
Lead slurry, with the reference of the tail slurry: $\log_{10}(61/24)/1.5812 = 0.4051/1.5812 = 0.26$ . With its own reference: $0.4051/\log_{10}(61/8) = 0.4051/0.8822 = 0.46$ .

Answer: The barrier between the perforated intervals is well bonded (BI 0.89). Above the upper interval the bond is partial (BI 0.51), not “84% bond” as the ratio of amplitudes would suggest. The lead slurry has BI 0.46 against its own reference, not 0.26.

What it means. BI 0.89 over all 16 m of the barrier is an argument that the perforated intervals are separated behind the casing. The choice of repair for the lower interval in the section on squeezing is based on this. An interval with BI 0.51 does not by itself tell us what is there: a channel, a microannulus or weak cement. We tell them apart from the wave pattern, from a log run under pressure and from the ultrasonic tool. If the cause is a channel, then in the simple model the cement fills about half of the circumference. A wrong reference is as dangerous as a wrong formula: against someone else's reference, lightweight cement looks almost uncemented. The data in this example are assumed for illustration.

The wave pattern and false readings

The wave pattern of the far receiver shows what we cannot see in the amplitude alone. The wave along the casing always arrives at the same time, because the casing is the same at all depths. On the VDL it appears as straight vertical bands. The wave along the rock arrives later, because rock is slower than steel. Its arrival time changes with the interval transit time of the rocks (the time sound needs to cross one metre of rock), so its bands become wavy. The direct wave through the mud arrives last. The rock wave reaches the receiver well when the cement is bonded to both the casing and the rock. It passes poorly through fluid in the annulus or through a gap, and it is lost against the background of the casing wave.

CalculatedA synthetic cement bond log, calculated with the lesson's model: the travel time and the amplitude at the near receiver, the bond index from the amplitude, and the VDL of the far receiver. Zones: 1 – free casing; 2 and 6 – bond to the casing and to the rock; 3 – bond to the casing only; 4 – partial bond (a channel); 5 – fast dense rock behind good cement. The data are assumed for illustration.

The synthetic log shows all the typical cases. In zone 1 the casing is free: the amplitude is 61 mV, and the VDL shows bright straight bands across the whole window. In zones 2 and 6 the cement is bonded to both the casing and the rock: the amplitude is about 1.6 mV, the casing bands are barely visible and the rock bands are bright. Zone 3 hardly differs from them in amplitude, but the VDL is empty. The cement holds on to the pipe, and a mudcake or water separates it from the rock, so flow along the wall of the hole is possible. In zone 4 we see both the casing and the rock, so the bond is partial. In zone 5 the amplitude is falsely high: dense fast rock sends its wave before the casing wave, and that wave falls into the measurement window. The travel time becomes shorter at the same time.

Signs on a cement bond log

State: Free casing
Amplitude, travel time: highest amplitude; constant time
VDL: bright straight casing bands across the whole window; kinks at the couplings

State: Bond to the casing and the rock
Amplitude, travel time: lowest amplitude
VDL: casing bands barely visible; rock bands bright and wavy

State: Bond to the casing only
Amplitude, travel time: lowest amplitude
VDL: almost empty: neither casing nor rock bands

State: Partial bond: a channel, weak cement
Amplitude, travel time: intermediate amplitude
VDL: both casing and rock are seen, both weakened

State: Microannulus
Amplitude, travel time: raised amplitude; falls under pressure
VDL: casing bands visible, rock bands clearly seen

State: Fast rock behind good cement
Amplitude, travel time: amplitude too high, time shortened
VDL: rock bands to the left of (earlier than) the casing bands

State: Tool pushed against the wall
Amplitude, travel time: amplitude too low, time shortened
VDL: casing bands blurred

State	Amplitude, travel time	VDL
Free casing	highest amplitude; constant time	bright straight casing bands across the whole window; kinks at the couplings
Bond to the casing and the rock	lowest amplitude	casing bands barely visible; rock bands bright and wavy
Bond to the casing only	lowest amplitude	almost empty: neither casing nor rock bands
Partial bond: a channel, weak cement	intermediate amplitude	both casing and rock are seen, both weakened
Microannulus	raised amplitude; falls under pressure	casing bands visible, rock bands clearly seen
Fast rock behind good cement	amplitude too high, time shortened	rock bands to the left of (earlier than) the casing bands
Tool pushed against the wall	amplitude too low, time shortened	casing bands blurred

A rock that is slower than the mud (loose sands, clays at shallow depth) gives no rock wave at all. An empty VDL then does not yet mean that the cement is not bonded to the rock.

How to interpret a cement bond log

1
Quality of the record
The amplitude in free casing matches the diameter and wall thickness of the pipe and the calibration of the tool. The travel time in free casing is constant. The repeat section (a second pass over the same interval) matches the main log.
2
Top of cement
This is the depth where the amplitude leaves the level of free casing and the rock appears on the VDL. We compare it with the design and with the temperature survey.
3
Intervals opposite the formations
Take the bond index against its own reference for each slurry portion, and use the VDL to see whether the cement is bonded to the rock. Pay special attention to the barriers between formations that must be kept apart and to the intervals to be perforated later.
4
False readings
A shortened travel time means an off-centre tool or fast rock. A high amplitude together with rock waves may mean a microannulus.
5
Doubtful intervals
Run the log again under pressure, use an ultrasonic tool and test with pressure. We decide on a repair according to what the defect connects.

A cement bond log is run once the cement has gained strength. This is usually no earlier than one or two days after the job, and later for lightweight and heavily retarded slurries. Liquid or weak cement weakens the wave only a little and gives a high amplitude, like free casing. A high amplitude together with clearly visible rock waves is a sign of a microannulus. This is a thin gap between the pipe and the set cement. It appears when the casing shrinks after the cement has set (the casing after cementing).

Worked exampleMicroannulus: how much the casing expands and shrinks

Given (assumed data)

Casing: 168.3 × 8.94 mm; steel: $E$ = 207,000 MPa (stiffness), $\alpha$ = 12·10⁻⁶ 1/°C (thermal expansion)
After the cement set: the casing was pressure-tested at 20 MPa; later it was cooled by 20 °C by pumping in cold fluid

Solution

The mean radius of the wall is $r = (168.3 - 8.94)/2 = 79.68$ mm. Growth of the radius with pressure: $r^2/(E\,t) = 79.68^2/(207{,}000\cdot 8.94) = 0.003431$ mm, that is 3.431 µm per 1 MPa.
The pressure test at 20 MPa expanded the casing by $20\cdot 3.431 = 68.6$ µm. If the set cement does not follow the pipe back, a gap of this order remains after the pressure is released.
Cooling by 20 °C shrinks the outer radius by $12\cdot10^{-6}\cdot 20\cdot 84.15 = 0.0202$ mm, that is by 20.2 µm.
To press the pipe back against the cement across such a gap, we need to add $20.2/3.431 = 5.9$ MPa of pressure inside.
If the cement holds the pipe along its length and does not let it shorten, two things change. The cooling shrinks the radius $1 + \nu = 1.3$ times more strongly. And the growth of the radius with pressure is multiplied by $1 - \nu^2 = 0.91$ ( $\nu = 0.3$ is Poisson's ratio of steel, which shows how much the steel narrows when it is stretched). The pressure needed is then $20.2\cdot 1.3/(3.431\cdot 0.91) = 8.4$ MPa.

Answer: Both a pressure test after the cement has set and cooling by 20 °C produce a microannulus of tens of micrometres. The gap from cooling is closed by an extra pressure of 5.9–8.4 MPa, depending on how firmly the cement holds the pipe along its length.

What it means. If the amplitude falls clearly when the log is run under pressure, the high amplitude came from a microannulus and not from a channel. A thin gap filled with liquid lets water through only weakly, but it passes gas tens of times better. The conductivity of a slit grows as the cube of its opening (the cubic law), and the viscosity of gas is tens of times lower. This is why the casing is pressure-tested at the bump, while the cement is still liquid (pressures during cementing), and not after the cement has set. The model is a thin-walled pipe. The last step shows how the answer changes with the way the cement holds the pipe along its length. So we run the log under pressure with a margin, at the pressure given in the programme. The data in this example are assumed for illustration.

Ultrasonic and other tools

An acoustic tool averages over the whole circumference. An ultrasonic scanner (a pulse-echo tool) looks in each direction separately. A rotating transducer (a sensor that sends and receives sound) sends a short pulse with a frequency of hundreds of kilohertz into the wall, at right angles to its surface. The pulse runs back and forth across the thickness of the wall, and the wall “rings” at the frequency at which its thickness equals half a wavelength. On each round trip some of the energy leaves the pipe and goes outward. The closer the acoustic impedance of the material behind the casing is to that of steel, the more energy leaves. (Acoustic impedance measures how hard it is for sound to move through a material: it is the density times the speed of sound.) So the ringing dies away faster behind cement than behind a liquid, and it lasts longest behind gas. The calculation is in the advanced section.

From the rate at which the ringing dies away, the tool finds the impedance of the material behind the pipe at every point of the circumference. It builds a map (an unrolled image of the casing wall) that shows solid (cement), liquid or gas. The map shows on which side the channel is and how wide it is. The time of the first reflection gives the inside radius of the pipe, and the ringing frequency gives the wall thickness. So the same measurement shows wear and corrosion of the casing. With the impedances taken in the lesson's example, the values behind the casing are: gas 0.07, water 1.5, lightweight cement 3.48 and class G cement 6.84 (in $10^6$ kg/(m²·s), called MRayl in English-language reports).

The method has limits too. The impedance of lightweight cement, and especially of foamed cement (cement made light with gas bubbles), can be close to that of heavy drilling mud, and the ultrasonic tool can hardly tell them apart. Heavy and gas-cut muds (muds with gas mixed into them) inside the casing weaken the signal strongly. A thin microannulus filled with liquid is noticed less by an ultrasonic tool than by an acoustic tool, and a microannulus filled with gas makes the cement look like gas. This is why tools that use flexural waves (bending waves) appeared. A slanted transmitter excites a flexural wave in the wall, and the wave gives up its energy into the annulus. Its attenuation separates solid from liquid better. The reflection from the wall of the hole also shows where the casing stands. Acoustic tools with sector (segmented) receivers measure the attenuation separately in several directions. In practice, the ultrasonic log is run together with the CBL and the VDL, and all are read together.

Squeeze cementing: purposes and mechanism

Squeeze cementing means forcing slurry into a place where there is not enough cement, or where a path for fluid must be closed. We use it to:

repair primary cementing: fill a channel or an uncemented interval, or seal the liner lap (where a liner, a short casing string hung inside the previous string, overlaps the casing above it) or the shoe
shut off a perforated interval that gives water or gas, or a formation that is no longer needed
isolate a formation before it is perforated, if there is no reliable barrier behind the casing
seal the casing at a leak or at a hole made by corrosion
close a lost-circulation zone while drilling
during abandonment, shut off a perforated formation together with a cement plug in the casing

Cement does not enter the pores of the rock. The average grain size of Portland cement is about 10–20 µm, and the large grains reach 100 µm. The pore channels of sandstone are much narrower. In the bundle-of-capillaries model their radius is $\sqrt{8k/\phi}$ , which is about 2 µm at a permeability of 100 mD and a porosity of 0.2. The grains jam at the entrance to the pores. Under the pressure difference, the slurry gives up water, the filtrate, to the rock. The solid material stays in the perforation tunnel (the channel that the charge punches into the rock) and on its wall as a dense cake of dehydrated cement. The cake grows until it blocks the hole, and then the pressure stops falling: the hole is closed. The slurry enters the rock whole only through fractures, caverns and channels that are a fraction of a millimetre wide or more. For thin channels and microannuli there are fine-grind cements with grains of a few micrometres.

This gives two ways to squeeze. In a squeeze without fracturing (low pressure), we keep the pressure on the formation below the fracture pressure (Fracture gradient and leak-off testing), and the slurry closes the perforations with a cake. The cake grows only where the slurry loses filtrate to the rock. So the perforation tunnels must be open before the squeeze. We wash out sand, scale and leftover mud, with acid if needed, and then check the injectivity: whether the formation takes fluid without fracturing at a small pump rate. In a squeeze with fracturing (high pressure), the cement is pumped into a fracture. We first open the fracture with water or brine (salt water), a fluid without solids, because drilling mud would plug the holes with its own cake. There is a price to pay. A fracture opens across the direction of the least stress (fracture orientation), and at depth it is usually vertical, so it can connect formations that must be kept apart. A lot of slurry is lost. And the fracture goes round the holes that are plugged with mud. So if the formation takes fluid without fracturing, we choose low pressure.

We choose the slurry to suit the formation (slurry testing). The fluid loss (how fast the slurry loses water to the rock) is especially important. If it is too high, the slurry dehydrates at the first holes and bridges them before it reaches the others. If it is too low, the cake grows slowly, the slurry stays liquid for a long time and goes further into the rock. The thickening time (the time before the slurry becomes too thick to pump) must cover pumping, the pauses and the washing out of the excess, with a margin. The volume is usually small: hundreds of litres or a few cubic metres. We need slurry at the holes, not in the formation.

The slurry is pumped either continuously (a running squeeze), at a low rate until the pressure rises to the final value, or with pauses (a hesitation squeeze). In the second way we pump small portions with pauses. During the pauses the cake has time to grow, and the pressure shows whether the formation still takes slurry. The pressure rises in steps. The squeeze ends when the pressure holds without falling for as long as the programme requires. Opposite each hole, a small mound of dehydrated cement stays inside the casing. The excess slurry is washed out.

SchematicThree ways to squeeze: through the wellhead without a packer, where the pressure acts on the whole casing and all the holes; with a retrievable packer above the interval, where the casing above the packer is relieved; and through a cement retainer with a bridge plug (a plug that seals the casing) below, where the slurry reaches only the isolated interval. The sketch is not to scale.

Through the wellhead, without a packer (bradenhead squeeze): the pipe takes the slurry down to the interval, the pipe is raised above the cement, the preventer is closed, and the slurry is forced in. It is simple, but the pressure acts on the whole casing, on the wellhead and on all the open holes. We cannot choose one interval out of several, and the pressure limit is the strength of the weakest part of the casing.
With a retrievable packer: a packer on the pipe seals the annulus above the interval, so the casing above it does not feel the squeeze pressure. The slurry is brought down to the packer through an open bypass (a passage that lets fluid go past the packer), and it is forced in with the bypass closed. In one trip the packer can be moved, and several intervals or leaks can be closed one after another. Everything that stands in the casing below the packer (about 18 L for each metre in 168.3 mm casing) goes into the formation ahead of the cement and mixes with it at the boundary. Cement that has risen to the packer can stick it. A length of pipe below the packer (a tail pipe) takes the slurry to the holes themselves, and the packer is left higher up.
Through a cement retainer: a drillable packer with a check valve (a valve that lets fluid through in one direction only). It is set on a wireline (a cable), with its depth matched to the casing collars, or on the pipe. Then the end of the pipe is pushed into it and the slurry is forced in. As soon as the pipe is pulled out, the valve closes. The cement then hardens under the retainer at the squeeze pressure, and the pipe above can be washed at once. The retainer is not recovered. If the interval is needed again, it is drilled out together with the cement.
Bridge plug: a mechanical plug with no passage for fluid, set below the squeeze interval. The slurry does not go down into the sump (the hole below the perforations) and does not reach the lower formations, and the plug serves as a support for it. Later it is drilled out or, if its design allows, recovered.
Through coiled tubing: a continuous pipe is run from a reel through the production tubing without pulling it out, and small volumes of slurry are placed accurately.

If water comes along a channel behind the casing from a neighbouring formation, squeezing the well's own perforations will not help. We open the channel with special holes above and below it and force the slurry into it. Or we circulate cement through the channel itself: down the pipe, out through the lower holes, up behind the casing and back into the casing through the upper holes. So before the repair we find out the path of the water, from the CBL, the temperature survey, noise logging (a log that listens to the sound of fluid moving) and the history of the well.

Squeeze design and checking the result

The main limit of a squeeze without fracturing is the pressure on the formation. We calculate it at rest. During the pauses the friction disappears, and the bottom of the hole feels the wellhead pressure plus the weight of the columns of fluids in the pipe. So the wellhead pressure can be as high as the limit at the bottom minus the pressure of those columns.

Wellhead pressure limit in a squeeze without fracturing

p_{\text{wh}}^{\max} = p_f - \Delta p_{\text{margin}} - 0.00981\sum_i \rho_i h_i

where:

$p_{\text{wh}}^{\max}$ highest allowed pressure in the pipe at the wellhead, at rest, MPa
$p_f$ fracture pressure at the depth of the holes, MPa
$\Delta p_{\text{margin}}$ margin that the programme leaves below the fracture pressure, MPa
$\rho_i,\ h_i$ densities of the fluids in the pipe and below the packer, down to the depth of the holes, and their vertical heights, g/cm³, m

Units: MPa; 0.00981 MPa is the pressure of a 1 m column of fluid with a density of 1 g/cm³

Holds when

The calculation is for rest. While we pump down the pipe, friction reduces the pressure at the bottom for the same wellhead pressure. So during pumping the pressure at the bottom stays below the limit, and in the pauses it just reaches it.
The fracture pressure is taken from tests of this formation or a neighbouring one, if there are any, and not from a forecast.
The heights are vertical: in a deviated well the columns are counted by the true vertical depth.

While heavy cement slurry is in the pipe, the wellhead limit is lower than with brine. A pressure that is safe in the injectivity test can fracture the formation once the pipe is full of cement. When the slurry goes below the packer or the retainer, the limit rises again.

Worked exampleSqueezing an interval that has started to produce water, through a retainer

Given (assumed data)

Well: casing 168.3 × 8.94 mm; upper perforated interval 2318–2326 m (oil), lower interval 2342–2350 m (now producing water); shale between them, BI 0.89 on the CBL
Task and equipment: shut off the lower interval with a squeeze without fracturing and keep the upper one; bridge plug at 2355 m, retainer at 2334 m; tubing 73 × 5.51 mm (inside diameter 61.98 mm)
Fluids: brine of 1.12 g/cm³ in the well; class G cement of 1.90 g/cm³ with a fluid-loss additive, 0.8 m³
Formation at 2346 m: fracture pressure 39.5 MPa (from a test in a neighbouring well), formation pressure 22.9 MPa; the programme keeps the bottomhole pressure 2.0 MPa below the fracture pressure
Injectivity: brine 0.15 m³/min at 6.0 MPa at the wellhead

Solution

Limit at the bottom: $39.5 - 2.0 = 37.5$ MPa. In the injectivity test the pressure at the bottom is at most $6.0 + 0.00981\cdot 1.12\cdot 2346 = 6.0 + 25.78 = 31.78$ MPa (friction in the tubing lowers it by a few tenths of a MPa). This is below the fracture pressure, so the formation takes brine without fracturing, and a low-pressure squeeze is possible.
While there is brine in the tubing, the wellhead limit is $37.5 - 25.78 = 11.72$ MPa.
All the slurry in the tubing fills a length of $0.8/0.003017 = 265.16$ m. The column down to 2346 m: $1.12\cdot(2334 - 265.16) + 1.90\cdot 265.16 + 1.12\cdot 12 = 2834.34$ . The wellhead limit is $37.5 - 0.00981\cdot 2834.34 = 9.70$ MPa, which is 2.02 MPa lower, and it stays so until the slurry leaves the retainer.
The slurry that leaves the retainer is heavier than brine. We assume that it sinks to the bridge plug and fills the casing from the bottom. The casing under the retainer holds $21\cdot 0.017771 = 0.373$ m³. When it is full, $0.8 - 0.373 = 0.427$ m³ (141.53 m) is left in the tubing, and cement stands opposite the holes. The wellhead limit is $37.5 - 0.00981\,[1.12\cdot(2334 - 141.53) + 1.90\cdot(141.53 + 12)] = 10.55$ MPa.
Pressure difference across the retainer at the limit of 37.5 MPa at the bottom: below it, at 2334 m, the pressure is $37.5 - 0.00981\cdot 1.90\cdot 12 = 37.5 - 0.22 = 37.28$ MPa. Above it there is brine: $0.00981\cdot 1.12\cdot 2334 = 25.64$ MPa. The difference is 11.64 MPa.
The upper interval at 2322 m stays under the pressure of the brine: $0.00981\cdot 1.12\cdot 2322 = 25.51$ MPa. In a squeeze through the wellhead without a retainer, it would have to bear up to $11.72 + 25.51 = 37.23$ MPa.

Answer: Keep the wellhead pressure no higher than 11.72 MPa while there is brine in the tubing, and no higher than 9.70 MPa while all the slurry is in it. The retainer takes a pressure difference of about 11.64 MPa and does not let the squeeze pressure reach the upper interval.

What it means. A wellhead pressure of 11.72 MPa is allowed in the injectivity test. Once slurry is in the tubing, the same wellhead pressure would put the pressure at the bottom 2.02 MPa above the limit, almost at the fracture pressure. We take more slurry than the casing under the retainer holds (0.373 m³). The rest goes into the perforation tunnels and serves as a reserve, and whatever is not needed is washed out by reverse circulation (pumping down the annulus and up the pipe) after the pipe is pulled off the retainer. The retainer is needed also because the upper interval is producing. In a squeeze through the wellhead it would be under the same pressure as the lower one and would take slurry. The choice of a retainer relies on the good barrier seen on the CBL. If water were coming behind the casing from the lower formation to the upper one, squeezing the holes would not stop it. The data in this example are assumed for illustration.

CalculatedThe calculated wellhead pressure limit at rest during the squeeze of the example. It falls while the slurry enters the tubing, stays level while the slurry is pushed down to the retainer, and rises when the slurry goes below the retainer. The horizontal line is the pressure in the injectivity test.

When the final pressure holds, we do one of two things. With a retainer, we pull the pipe out of it and the valve closes. With a packer, we let the pressure off and release the packer. The excess slurry is washed out of the pipe by reverse circulation, and we wait for the cement to set. If the interval is needed again, the cement in the casing is drilled out. Then the squeeze is checked by pressure in both directions. One is a pressure test from inside, at a pressure below the fracture pressure. The other is an inflow test (lowering the level in the well). In the inflow test we lower the pressure in the casing below the formation pressure. There are three ways: swabbing (pulling fluid out of the well with a cup-type tool), replacing the fluid with a lighter one, or using a drill stem test tool (a tool run on the pipe to test a formation). Then we watch whether the level or the pressure rises.

Cement plugs

A cement plug closes the wellbore, cased or open, over a chosen length. We set a plug to abandon or suspend a well or one formation, to close a lost-circulation zone, to start a sidetrack, which is a new branch drilled out of the old hole (sidetracking from a cased well), or to isolate intervals during a test.

A plug is a small volume of slurry in a large volume of another fluid, and that is what makes it hard. A few hundred litres of mixing at the boundaries with the spacer and with the mud make up a noticeable share of the plug. So we make the plug longer than needed, separate it from the drilling mud with spacers, and get the mud in the well ready before the job. Cement slurry is heavier than drilling mud, and a heavy column standing on a lighter fluid is unstable. The cement tends to sink, and the mud under it tends to rise in tongues, especially in a deviated hole. So we set a plug on a support: the bottom of the hole, a bridge plug, a plug set earlier, or a viscous pill (a small batch of fluid) thickened so that it holds the cement.

Balanced plug: the main method. Spacer, cement and spacer again are pumped down open-ended pipe and displaced so that the levels in the pipe and in the annulus become equal. Then the pipe is pulled out slowly, and the cement stays in the wellbore.
Dump bailer: a container run on a wireline (a cable) and emptied right on a bridge plug. This is how we set a short plug in the casing, a “cap” of cement on the bridge plug. The cable sets the depth, and the slurry does not travel through a column of another fluid. A bailer carries only a little slurry per run, and the slurry must stay liquid while the bailer is run in.
With separating plugs: the cement is pumped down the pipe between two plugs, and the plugs are caught at the end of the pipe. This gives less mixing on the way down.
Through coiled tubing: the plug is set in a producing well through the production tubing, without pulling it out.

Balanced plug: heights and volumes

h_{\text{cem}} = \frac{V_{\text{cem}}}{A_{\text{ann}} + A_{\text{pipe}}}, \qquad V_{\text{sp2}} = V_{\text{sp1}}\,\frac{A_{\text{pipe}}}{A_{\text{ann}}}, \qquad V_{\text{disp}} = A_{\text{pipe}}\,(H - h_{\text{cem}} - h_{\text{sp}}), \qquad L_{\text{plug}} = \frac{V_{\text{cem}}}{A_{\text{hole}}}

where:

$V_{\text{cem}}$ volume of cement slurry for the plug, m³
$h_{\text{cem}}$ height of the cement while the pipe is in the hole; it is the same in the pipe and in the annulus, m
$A_{\text{ann}},\ A_{\text{pipe}},\ A_{\text{hole}}$ cross-section areas of the annulus (the hole minus the outside diameter of the pipe), of the channel inside the pipe, and of the hole without the pipe, m²
$V_{\text{sp1}},\ V_{\text{sp2}}$ volume of the spacer ahead of the cement (it goes into the annulus) and behind it (it stays in the pipe), m³
$h_{\text{sp}}$ height of the spacer: $V_{\text{sp1}}/A_{\text{ann}}$ , the same in the annulus and in the pipe, m
$V_{\text{disp}},\ H$ volume of displacement and depth of the end of the pipe (the bottom of the plug), m³, m
$L_{\text{plug}}$ length of the plug after the pipe is pulled out, m

Units: SI: m, m², m³

Holds when

The fluids do not mix. The same fluids stand at the same height in the pipe and in the annulus. Then the pressures at the end of the pipe are equal, and there is no flow.
The cross-section of the hole is constant (taken from the caliper log, a log that measures the diameter of the hole). The tool joints are ignored. They reduce the annulus slightly, so with the pipe in the hole the cement stands a little higher than calculated. The deviation of the hole is ignored.
The pipe drains as it is pulled out, and all the cement stays in the wellbore.

While the pipe is in the hole, it takes up part of the cross-section, so the cement stands higher than the plug will be: $h_{\text{cem}} > L_{\text{plug}}$ . The spacer behind the cement is sized so that it takes the same height in the pipe as the spacer ahead of the cement takes in the annulus.

Worked exampleA balanced plug in an open hole

Given (assumed data)

Plug: from the bottom of the hole at 2735 m up to 2668 m, opposite water-bearing formation C (see the abandonment example below)
Hole: bit 215.9 mm, 228 mm on the caliper log; drilling mud 1.18 g/cm³
Pipe: drill pipe 127 × 9.19 mm (inside diameter 108.62 mm), open end at the bottom of the hole
Fluids: water ahead, 1.6 m³; class G cement of 1.90 g/cm³, slurry yield 0.757 m³/t; displacement with drilling mud, 0.2 m³ short

Solution

Areas: the hole $\frac{\pi}{4}\cdot 0.228^2 = 0.040828$ m², the channel inside the pipe $\frac{\pi}{4}\cdot 0.10862^2 = 0.009266$ m², the annulus $\frac{\pi}{4}(0.228^2 - 0.127^2) = 0.028160$ m².
Slurry for 67 m: $67\cdot 0.040828 = 2.735$ m³, which takes $2.735/0.757 = 3.61$ t of cement.
While the pipe is in the hole: $h_{\text{cem}} = 2.735/(0.028160 + 0.009266) = 73.08$ m, and the top of the cement is at $2735 - 73.08 = 2661.9$ m.
The water ahead takes $1.6/0.028160 = 56.82$ m in the annulus. Behind the cement we put the same number of metres in the pipe: $56.82\cdot 0.009266 = 0.526$ m³.
The displacement is the volume of the pipe from the wellhead down to the top of the water in it: $0.009266\cdot(2735 - 73.08 - 56.82) = 24.14$ m³. We pump 23.94 m³, which is 0.2 m³ short. The cement and the water in the pipe then stay $0.2/0.009266 = 21.6$ m above the equilibrium level, and in the annulus they stay $0.2/0.028160 = 7.1$ m below it.
After the pipe is pulled out, the cement fills the whole hole: $2.735/0.040828 = 67.0$ m, from 2668 to 2735 m.

Answer: We need 2.735 m³ of slurry (3.61 t of cement), 0.526 m³ of water behind the cement, and 23.94 m³ of drilling mud as displacement. In equilibrium with the pipe in the hole, the top of the cement is at 2661.9 m, and after the pipe is pulled out it is at 2668 m.

What it means. While the pipe is in the hole, the cement stands 6.08 m higher than the future plug, because the pipe takes up part of the cross-section. Pumping extra displacement would first push cement out of the pipe into the annulus and upset the balance. Pumping more than 0.677 m³, which is all the cement in the pipe, would push the water, and after it the drilling mud, out from under the end of the pipe, below the cement. They would then float up through the plug. Pumping short leaves a small excess in the pipe, and it runs down onto the plug as the pipe is pulled out. The water ahead reduces the pressure below its top by 0.10 MPa compared with mud, and the cement in place of mud adds 0.52 MPa at the bottom of the plug. So the overbalance on the formations is kept (the pressure in the well stays above the formation pressure). The plug stands on the bottom of the hole, which is a firm support. If the hole is narrower than the caliper log shows, the cement will rise higher than calculated and the balance will shift. So we take a margin in the length of the plug and check its top afterwards. The data in this example are assumed for illustration.

CalculatedThe balanced plug of the example. On the left the pipe is still in the hole: the cement and the water stand at equal heights in the pipe and in the annulus. On the right the pipe has been pulled out: the cement fills the whole cross-section, and the plug is shorter. The heights are calculated from the data of the example, and the width is not to scale.

The pipe is pulled out slowly so as not to mix the cement, up to a depth well above the top of the plug. Then it is flushed by circulating fluid down the pipe (forward circulation) or down the annulus (reverse circulation), because excess cement must not stay in the pipe and harden there. After the cement has set, we find the top of the plug by tagging it: we lower drill pipe or tubing onto the plug and set part of its weight on it. If the top of the plug is lower than planned, or the plug is soft, it is set again. A plug in the casing is also pressure-tested.

A plug for a sidetrack must be harder than the surrounding rock. Otherwise the bit goes down the old hole. We use a slurry with a lower water–cement ratio and a plasticiser, which is denser and stronger than usual (density and yield), we make the plug longer, and before the sidetrack we drill off its soft top. In a cased hole the sidetrack is more often started from a whipstock (a steel wedge that pushes the bit sideways). Under the industrial safety rules of Kazakhstan, a plug with a whipstock for a sidetrack in the casing is tested by setting down the weight of the drilling tool, not above the load allowed on set cement. It is also pressure-tested, together with the casing and the installed blowout preventer, at the maximum pressure that can arise when a kick (an unwanted flow of formation fluid into the well) occurs and while it is being brought under control (order No. 355, paragraph 125).

Abandonment and suspension of wells

Abandonment of a well (plug and abandonment, P&A) is the final end of its work. The isolation must work for ever, with no supervision and no repair. In Kazakhstan the official word is “liquidation”. Suspension (the official word is “conservation”) is a temporary stop. The well is filled with fluid and the wellhead is kept, so that the well can be returned to work later. The requirements for abandonment are stricter. A plug that nobody will check for a hundred years must be set where the rock itself will hold it.

In international practice (for example, in the Norwegian well integrity standard NORSOK D-010), the final barrier is described like this. It closes the whole cross-section of the well, both the wellbore and the annulus behind the casing, “from rock to rock”. It stands opposite impermeable rock that can withstand the pressure that may build up under it. For each formation from which flow is possible, a primary and a backup barrier are set, and one more near the surface. A plug inside the casing is useless if there is no cement behind the casing at the same depth. So the cement behind the casing is confirmed by logging. If it is missing, a section of the casing is cut out or perforated, washed out, and the annulus is cemented. The plugging material must withstand formation waters, hydrogen sulfide and carbon dioxide.

In Kazakhstan the procedure is set by the Rules for conservation and liquidation in the exploration and production of hydrocarbons (order of the Minister of Energy of 22 May 2018 No. 200). The technical requirements for the wells are in annex 8 to the rules. Ordinary wells are abandoned and suspended under a plan. A plan is made for each well and approved by the subsoil user (the company that holds the licence). Among other things, the plan contains the cementing calculations and the requirements for the plugging materials. Complex wells are those with hydrogen sulfide in the reservoir fluid at 3.5% or more, those with an abnormality coefficient (the ratio of pore pressure to the pressure of fresh water) of 1.5 or more, and those deeper than 5000 m on land. Complex wells, and wells at sea and in inland waters, are suspended and abandoned under a project of a licensed design organisation. The project must have passed the expert reviews. Before the work, the downhole equipment is taken out and the hole is cleaned out down to the artificial bottom (the current bottom of the hole, for example the top of a plug set earlier).

Ordinary well with a production casing: the plugs cover the perforated intervals completely and another 20 m below and above. The same applies to leaking intervals, to stage collars (the tools used in stage cementing) and to the joints between sections of the casing strings. At the shoe of the last casing string, the plug runs from 20 m below to 50 m above the shoe (paragraph 28).
Ordinary well without a production casing: a plug opposite each formation that holds oil, gas or water: the thickness of the formation plus 20 m above its top and 20 m below its base. Above the top of the highest formation, the plug is at least 50 m high (paragraph 29). If there are no such formations, there is a plug of at least 50 m at the shoe of the last casing string (paragraph 30).
Complex or offshore well: a productive formation is covered by a plug over its whole thickness and 100 m above its top. Without a production casing, there is also a plug of at least 100 m at the shoe of the last intermediate casing string. The docking (connecting) devices in the last casing string run are covered by a plug from 50 m below to 50 m above (paragraph 31).
Plugging material: for plugs it must resist corrosion and meet the requirements that the drilling design sets for cementing casing strings across formations with hydrogen sulfide (paragraph 32).
Checking the plugs: the presence of the plugs is checked by setting down the weight of drill pipe or tubing, with a force not higher than the allowed load per unit area on set cement. A plug at the shoe of the last intermediate casing is also pressure-tested (paragraph 33).
Wellhead: on the wellhead of an abandoned well there is a reinforced concrete block of 1 × 1 × 1 m with a plate showing the number and coordinates of the well, the field, the subsoil user and the date of abandonment (paragraph 34). The wellhead of a complex well on land is fitted with a casing head and a high-pressure valve in a corrosion-resistant design, with outlets to monitor the pressure in the tubing and between the casing strings, and with a metal plate (paragraph 35). At sea the casing strings are cut below the sea floor, the top of the well is filled with cement up to the level of the sea floor and covered with concrete slabs (paragraphs 36–37).

During suspension, the well is filled with a fluid that does not corrode the casing, protects the reservoir and puts a back pressure on the formation. The upper part is filled with a non-freezing fluid. The plan decides whether a cement plug is needed above the perforations. In a complex or offshore well it is mandatory, at least 100 m high (paragraphs 16–17). For an offshore well, paragraph 20 describes the plug differently, “50 metres high, 20 metres below the perforated interval and 50 metres above”, and this does not agree with paragraph 17 (suspension of an offshore well). A contractor holding a licence for this kind of work does the abandonment. The acceptance report is signed by a commission set up by an order of the subsoil user. For complex and offshore wells the industrial safety authority takes part. All the documents are kept in the well file. The paragraph numbers follow annex 8 to the rules, and the text is as of October 2026.

Worked examplePlugs for abandoning an exploration well

Given (assumed data)

Well: an exploration well on land, 2735 m deep; no hydrogen sulfide, formation pressures close to hydrostatic: an ordinary well, abandoned under a plan
Well design: intermediate casing 244.5 × 11.05 mm (inside diameter 222.4 mm) down to 1660 m, open hole of 215.9 mm below it; no production casing was run
Formations that hold fluids: A 1702–1716 m (water), B 2194–2207 m (oil with water, a flow too small to be commercial), C 2688–2709 m (water)
Hole diameter on the caliper log: 236 mm opposite A, 232 mm opposite B, 228 mm opposite C
Cement: class G, 1.90 g/cm³, slurry yield 0.757 m³/t

Solution

The rule for a well without a production casing: the plug covers the formation and another 20 m above its top and 20 m below its base, and above the top of the highest formation at least 50 m.
Formation C: from $2688 - 20 = 2668$ to $2709 + 20 = 2729$ m. There are 6 m left to the bottom of the hole, so the plug is set from the bottom: 2668–2735 m, 67 m long, $67\cdot 0.040828 = 2.735$ m³. This is the plug calculated in the example above.
Formation B: from $2194 - 20 = 2174$ to $2207 + 20 = 2227$ m, 53 m long, $53\cdot 0.042273 = 2.240$ m³. Below it, down to plug C, there are 441 m of mud, so the plug needs a support: a viscous pill or a bridge plug.
Formation A is the highest: from $1702 - 50 = 1652$ to $1716 + 20 = 1736$ m, 84 m long. The plug enters the casing by 8 m and covers its shoe: $8\cdot 0.038847 + 76\cdot 0.043744 = 3.635$ m³. Below it, down to plug B, there are 438 m of mud, so it needs a support too.
In total $2.735 + 2.240 + 3.635 = 8.610$ m³ of slurry, which is $8.610/0.757 = 11.37$ t of cement.
Checking: the top of each plug is found by setting down the weight of drill pipe. Plug A stands at the shoe of the last intermediate casing, so it is also pressure-tested.

Answer: Three plugs: 2668–2735, 2174–2227 and 1652–1736 m; 8.610 m³ of slurry, 11.37 t of cement. Plug A is checked by setting down weight and by a pressure test.

What it means. The rules set a minimum. The plan may add plugs and length. Plugs B and C stand in the open hole and close it “from rock to rock”. Plug A at the top also relies on the cement behind the intermediate casing at the shoe. So both its pressure test and the cement bond log of that casing are important. The plugs are set from the bottom up, and each is checked before the next one is set. We bring cement with a reserve for losses when mixing the slurry and for mixing at the boundaries. The casing above plug A is filled with drilling mud, and the wellhead gets a concrete block with a plate. The data in this example are assumed for illustration.

CalculatedThe plugs for abandoning the exploration well of the example: intermediate casing down to 1660 m, open hole down to 2735 m, the formations A, B and C that hold fluids, and the plugs with the margins required by the rules: 20 m below the base and above the top, and 50 m above the top of the highest formation. The depths are taken from the data of the example, and the width is not to scale.

Terms and symbols

Cement bond log (CBL): A record of the wave that has travelled along the casing: the amplitude and the travel time at the near receiver, and the wave pattern at the far one.
Variable density log (VDL): An image of the wave patterns against depth. Positive half-cycles are dark; straight bands are the casing, and wavy bands are the rock.
Bond index (BI): The ratio of the added attenuation in an interval to the attenuation in an interval of full bond of the same cement.
Microannulus: A thin gap between the casing and the set cement after the casing has shrunk. It makes the CBL amplitude too high and closes under pressure.
Ultrasonic scanner: A tool that measures the impedance of the material behind the casing round the circumference from the decay of the ringing of the pipe wall.
Squeeze cementing: A job that forces slurry into perforations, channels or leaks. Without fracturing the rock, the slurry dehydrates and leaves a cake in the holes.
Cement retainer: A drillable packer with a valve. The slurry is forced through it, and after the pipe is pulled out the valve holds the cement in place.
Bridge plug: A plug that closes the casing and holds pressure from above and from below. It is a support for cement and it separates intervals.
Cement plug: A plug of set cement that closes the wellbore over a chosen length. A balanced plug is set with equal levels in the pipe and in the annulus.
Inflow test: A test of tightness made by lowering the pressure in the well below the formation pressure. It loads the holes and plugs in the way the formation will load them.
Abandonment and suspension: The final and the temporary end of the work of a well. At abandonment, the isolation must work for ever.

Common mistakes and misconceptions

Wrong

“The amplitude is 16% of that in free casing, so 84% of the circumference is bonded.”

In fact

Cement weakens the wave exponentially, so the bonded share is calculated from the attenuation, through the logarithm of the amplitude. In the lesson's example such an amplitude gives a bond index of about 0.5.

Wrong

“The amplitude is at its lowest, so the formations are isolated.”

In fact

A low amplitude is also given by a tool pushed against the wall (the travel time is then shortened), and by cement that is bonded to the pipe only (there are no rock waves on the VDL). Isolation is confirmed by a combination of the amplitude, the VDL and tests.

Wrong

“The amplitude is high, so there is no cement and a repair is needed.”

In fact

First we rule out a microannulus (a log run under pressure), fast rock (a shortened time, rock arrivals ahead of the casing) and lightweight cement (its own reference for full bond). A repair based on a false reading costs money and can damage good cement.

Wrong

“The higher the squeeze pressure, the better the cement enters the formation.”

In fact

Without fracturing, cement does not enter the pores at all. It dehydrates and closes the holes with a cake. Above the fracture pressure the formation splits, the fracture runs across the least stress, often vertically, and the slurry is lost in it or connects formations.

Wrong

“The plug was pumped as calculated, so it is in place.”

In fact

The plug could have sunk into a lighter fluid or mixed with the spacer or with the mud at the boundaries. We find its top by tagging with the pipe after it has set, and a plug in the casing is also pressure-tested.

In depth: advanced detail

Travel time: an off-centre tool and fast rocks

The wave along the casing is the same head wave (a wave that runs along the boundary between two media) as in an open hole (travel time of the head wave). Here the “rock” is the wall of the pipe. If the tool is displaced by $e$ from the axis, the gap is smaller on one side and larger on the other, and the arrivals from the different sides come with different delays. The receiver adds them up. For a narrow-band signal (a signal with almost a single frequency), the sum is weaker than for a centred tool.

Amplitude of an off-centre tool

\frac{A_e}{A} = J_0\!\left(2\pi f\,\tau\right), \qquad \tau = 2e\,\Delta t_m\cos\theta_c, \qquad \sin\theta_c = \frac{\Delta t_{\text{cas}}}{\Delta t_m}

where:

$A_e,\ A$ amplitude of the casing wave with an off-centre tool and with a centred tool, mV
$J_0$ Bessel function of the first kind of order zero, —
$f$ frequency of the signal, Hz
$\tau$ how much earlier the wave arrives from the near side (and later from the far side), s
$e$ displacement of the tool from the axis of the casing, m
$\Delta t_m,\ \Delta t_{\text{cas}}$ interval transit times of the mud and of the casing wave, µs/m

Units: SI; $\tau$ in the formula $2\pi f\tau$ is in seconds

Holds when

The delay round the circumference varies as $\tau\cos\varphi$ (the displacement is small compared with the radius of the casing), and the arrivals from all sides are equally strong.
The signal is narrow-band: a sine wave of frequency $f$ . For a short pulse the weakening is of the same order.
The ray approximation is used, as in the formula for the travel time of the head wave.

The sum of cosines with phases $2\pi f\tau\cos\varphi$ , averaged round the circumference, is $J_0$ . When $\tau$ reaches a quarter of the period, the amplitude falls by half.

Worked exampleTravel time, tool displacement and dense rock

Given (assumed data)

Casing and tool: 168.3 × 8.94 mm (inside diameter 150.42 mm); an 86 mm tool in the centre, so the gap is 32.21 mm
Interval transit times: casing 187 µs/m, mud 650 µs/m, set cement 280 µs/m; the cement sheath in a 215.9 mm hole is 23.8 mm thick
Signal and spacing: 20 kHz (period 50 µs); near receiver at 0.9144 m (3 ft); free casing 61 mV, reference for full bond 1.6 mV

Solution

Angle: $\cos\theta_c = \sqrt{1 - (187/650)^2} = 0.95772$ . Time along the casing: $0.9144\cdot 187 + 2\cdot 0.03221\cdot 650\cdot 0.95772 = 171.0 + 40.1 = 211.1$ µs.
Displacement of 6 mm: $\tau = 2\cdot 0.006\cdot 650\cdot 0.95772 = 7.47$ µs, $2\pi\cdot 20{,}000\cdot 7.47\cdot 10^{-6} = 0.939$ , and the amplitude is multiplied by $J_0(0.939) = 0.79$ . Displacement of 10 mm: $\tau = 12.45$ µs, $J_0(1.565) = 0.48$ .
Free casing with a 10 mm displacement will show $61\cdot 0.48 = 29.3$ mV and a bond index $\log_{10}(61/29.3)/1.5812 = 0.20$ , with no cement at all. The travel time will become shorter by 12.45 µs.
Dense rock of 150 µs/m behind good cement: $0.9144\cdot 150 + 2\cdot 0.03221\cdot 650\cdot 0.97301 + 2\cdot 0.0238\cdot 280\cdot 0.84440 = 137.2 + 40.7 + 11.3 = 189.2$ µs. The cosines are for $\sin\theta = 150/650$ and $150/280$ . The rock wave arrives 21.9 µs earlier than the casing wave.
The rock wave arrives ahead of the casing wave if its interval transit time is less than 175 µs/m (53 µs/ft), for this gap and this cement thickness.

Answer: A displacement of 10 mm cuts the amplitude almost in half and shortens the time by 12.45 µs, so free casing looks partly cemented (BI 0.20). Dense rock arrives 21.9 µs ahead of the casing wave and makes the amplitude too high when the cement is good.

What it means. This gives a rule for checking the record. A shortened time together with a too-low amplitude is a sign of an off-centre tool. A shortened time together with a too-high amplitude is a sign of fast rock. That is why CBL tools are run on centralisers, and the travel time in free casing is recorded and compared with the calculated one. Dense limestones, dolomites and anhydrites are usually faster than the casing, and sandstones and shales are slower. The model uses rays and a narrow band, and it ignores the casing wall on the path of the rock wave. At 150 µs/m the wall would add about 1.4 µs and reduce the lead by the same amount. The threshold does not change, because at the threshold the rock is slower than the compressional wave in steel (about 170 µs/m), and there is no ray delay in the wall. So the numbers are estimates of the order of magnitude. The data in this example are assumed for illustration.

Why the ringing of the wall dies away faster behind cement

An ultrasonic pulse that enters the wall of the pipe runs back and forth between its inner and outer surfaces. At each surface, part of the wave is reflected back into the steel, and part goes out into the mud inside the casing or into the material behind it. The share of the amplitude that is reflected is the reflection coefficient, the same as in seismic exploration (impedance and reflection).

Decay of the ringing of the casing wall

D = 20\,\log_{10}\left|R_{\text{in}}\,R_{\text{out}}\right|, \qquad R = \frac{Z_2 - Z_1}{Z_2 + Z_1}, \qquad f_0 = \frac{v_{\text{steel}}}{2\,t}

where:

$D$ change of the amplitude of the ringing over one round trip across the wall and back, dB
$R_{\text{in}},\ R_{\text{out}}$ reflection coefficients at the inner surface (steel to mud) and at the outer surface (steel to the material behind the casing), —
$Z_1,\ Z_2$ acoustic impedances of steel and of the material on the other side of the surface, $10^6$ kg/(m²·s)
$f_0$ frequency of the ringing: the thickness resonance of the wall, Hz
$v_{\text{steel}},\ t$ speed of the compressional wave in steel (about 5900 m/s) and the wall thickness, m/s, m

Units: SI; impedance is density times speed

Holds when

A plane wave at normal incidence. The attenuation in the steel itself and the spreading of the beam are ignored. They make the decay faster behind any material, so real times are shorter than calculated, and the tool finds the difference between materials by calibration.
Behind the casing there is a layer that is thick compared with the wavelength. A thin layer or a microannulus changes the picture.
Real processing fits a model to the whole record, with the transducer and the mud taken into account. The formula explains the principle.

The closer the impedance of the material behind the casing is to that of steel, the smaller $|R_{\text{out}}|$ is and the faster the ringing dies away. So we recover the impedance behind the pipe from the rate of decay.

Worked exampleRinging of the wall behind cement and behind water

Given (assumed data)

Casing wall: 8.94 mm; steel of 7.85 g/cm³ and 5.9 km/s, impedance 46.315
Inside the casing: brine of 1.12 g/cm³ and 1.55 km/s, impedance 1.736
Behind the casing: class G cement (1.90 g/cm³, 3.6 km/s), impedance 6.84, or water (1.0 g/cm³, 1.5 km/s), impedance 1.5; impedances in $10^6$ kg/(m²·s)

Solution

Round trip across the wall and back: $2\cdot 8.94\cdot 10^{-3}\ \text{m}/5900\ \text{m/s} = 3.031\ \mu\text{s}$ . The ringing frequency is $1/(3.031\ \mu\text{s}) = 330\ \text{kHz}$ .
Inner surface: $|R_{\text{in}}| = (46.315 - 1.736)/(46.315 + 1.736) = 0.9277$ .
Behind cement: $|R_{\text{out}}| = (46.315 - 6.84)/(46.315 + 6.84) = 0.7426$ . Per round trip $0.9277\cdot 0.7426 = 0.6889$ , that is $20\log_{10} 0.6889 = -3.237$ dB. The ringing weakens by 20 dB in $20/3.237\cdot 3.031 = 19$ µs.
Behind water: $|R_{\text{out}}| = 0.9373$ . Per round trip $0.9277\cdot 0.9373 = 0.8695$ ( $-1.215$ dB), and the 20 dB drop takes $20/1.215\cdot 3.031 = 50$ µs.

Answer: The wall rings at 330 kHz. Behind class G cement the ringing dies down by 20 dB in about 19 µs, and behind water in 50 µs, which is about 2.6 times longer.

What it means. The ringing frequency sets the working range of the tool: walls from 5 to 15 mm ring at about 590 to 200 kHz. The difference in decay behind cement and behind water is large, but behind lightweight cement it is smaller (see the table below), because its impedance is closer to that of liquids. The data in this example are assumed for illustration.

Ringing of an 8.94 mm wall with brine inside the casing

Behind the casing: Gas
$Z$ , $10^6$ kg/(m²·s): 0.0675
$|R_{\text{out}}|$: 0.9971
Per round trip, dB: −0.677
Time to −20 dB, µs: 90

Behind the casing: Water
$Z$ , $10^6$ kg/(m²·s): 1.5
$|R_{\text{out}}|$: 0.9373
Per round trip, dB: −1.215
Time to −20 dB, µs: 50

Behind the casing: Lightweight cement
$Z$ , $10^6$ kg/(m²·s): 3.48
$|R_{\text{out}}|$: 0.8602
Per round trip, dB: −1.960
Time to −20 dB, µs: 31

Behind the casing: Class G cement
$Z$ , $10^6$ kg/(m²·s): 6.84
$|R_{\text{out}}|$: 0.7426
Per round trip, dB: −3.237
Time to −20 dB, µs: 19

Behind the casing	$Z$ , $10^6$ kg/(m²·s)	$\|R_{\text{out}}\|$	Per round trip, dB	Time to −20 dB, µs
Gas	0.0675	0.9971	−0.677	90
Water	1.5	0.9373	−1.215	50
Lightweight cement	3.48	0.8602	−1.960	31
Class G cement	6.84	0.7426	−3.237	19

The densities and speeds are as in the example above, plus gas of 0.15 g/cm³ and 0.45 km/s and lightweight cement of 1.45 g/cm³ and 2.4 km/s. $|R_{\text{in}}|$ = 0.9277, and the round trip takes 3.031 µs. Lightweight cement makes the ringing die away only one and a half times faster than water does. That is why it is hard to work with.

Connections to other topics

Primary cementing — the cementing that we check and repair: centralisation, displacement of the mud, volumes, pressures, and the pressure test of the casing at the bump.
Oilwell cements and additives — recipes for squeezes and plugs: fluid loss, thickening time, density and yield of the slurry, heat of hydration.
Sonic log — the head wave, receivers at 0.9 and 1.5 m, and the interval transit time of rocks: the basis of acoustic cement logging.
Seismic waves, acoustic impedance and reflections — acoustic impedance and the reflection coefficient: the basis of ultrasonic tools.
Well control: primary barrier and kick indicators — well barriers; the cement sheath and plugs as part of them.
Well completion: opening the reservoir and the tubing string — the integrity of a producing well as a system of barriers, and monitoring of the pressure between casing strings.
Fracture gradient and leak-off testing — the fracture pressure is the limit in a squeeze; the test below the shoe.
Rock mechanics for hydraulic fracturing — why the fracture in a squeeze with fracturing opens across the least stress.
Casing design — what happens to the casing after cementing when the pressure and temperature change.
Deflection tools and downhole motors — the whipstock and the motor with a bent housing, which are used to start a sidetrack.
Abnormal formation pressures: origin, prediction and recognition — the abnormality coefficient, which the rules of Kazakhstan use to define complex wells.

Check yourself

1
In free casing the amplitude is 58 mV, in an interval of full bond of the same cement it is 2.0 mV, and in the interval under study it is 5.8 mV. What is the bond index?
1. A.About 0.32
2. B.About 0.10
3. C.About 0.90
4. D.About 0.68
Show the answer
Correct answer: D. About 0.68
$\text{BI} = \log_{10}\!(58/5.8)/\log_{10}\!(58/2.0) = 1/1.462 \approx 0.68$ . The ratio of amplitudes, 5.8/58 = 0.10, is not the bonded share, and “1 − 0.10 = 0.90” is not either. Cement weakens the wave exponentially, so the bonded share comes from logarithms of the amplitudes.
2
In an interval the CBL amplitude is at its lowest, the travel time is normal, and the VDL is almost empty, with neither casing bands nor rock waves. What is this most likely?
1. A.Free casing
2. B.Good bond to the casing and to the rock
3. C.A microannulus
4. D.Cement is bonded to the casing but not to the rock
Show the answer
Correct answer: D. Cement is bonded to the casing but not to the rock
A weak casing wave means a good bond to the pipe, and the absence of rock waves means that the energy does not reach the rock: there is a mudcake or water between the cement and the wall of the hole. A rock that is slower than the mud gives the same picture, so we check against the geological section. In free casing the amplitude is the highest. With a good bond to both surfaces, bright wavy rock bands are visible. A microannulus gives an intermediate amplitude.
3
A temperature survey 12 hours after cementing showed the top of the cement at 1450 m instead of the planned 1150 m, although the slurry was pumped according to the programme. What two causes are possible, how can we tell them apart, and what should be done next?
Show the answer
Either the hole is wider than assumed (washouts), or part of the slurry went into the formation (lost circulation). We tell them apart from the record of the job and from the caliper log. With lost circulation, the returns from the annulus were smaller than the pumping rate, and the pressure at the end of the displacement was lower than calculated. Then, when the cement has gained strength, we run a CBL to see whether the formations that must be kept apart are covered and whether the design requirement for the height of the cement is met. If not, the repair is done through special holes: a squeeze, or circulating cement through two rows of holes.
4
A plug is set through 88.9 × 6.45 mm tubing (inside diameter 76.0 mm) in a casing with an inside diameter of 150.4 mm. 1.2 m³ of water was pumped ahead of the cement. How much water must be pumped behind the cement, and why?
Show the answer
The annulus is $\frac{\pi}{4}(0.1504^2 - 0.0889^2) = 0.011559$ m² and the channel inside the tubing is $\frac{\pi}{4}\cdot 0.076^2 = 0.004536$ m². The water ahead takes $1.2/0.011559 = 103.8$ m in the annulus. Behind the cement the same height is needed in the tubing: $103.8\cdot 0.004536 \approx 0.47$ m³. Then the same fluids stand at the same height in the pipe and in the annulus, the pressures at the end of the pipe are equal, and nothing flows after the pumps stop.
5
After the lower interval was squeezed, the casing was pressure-tested and the pressure holds. Why is this not enough?
1. A.A pressure test must not be done after a squeeze
2. B.The pressure test checks only the wellhead equipment
3. C.The test pressure is always lower than the formation pressure
4. D.Dehydrated cement in the holes can hold pressure from inside, like a stopper pushed in, but fail under the formation pressure from outside, so an inflow test is needed
Show the answer
Correct answer: D. Dehydrated cement in the holes can hold pressure from inside, like a stopper pushed in, but fail under the formation pressure from outside, so an inflow test is needed
Pressure from the casing presses the cement cake against the hole, and the formation pressure pushes it out. The inflow test, which lowers the pressure in the well below the formation pressure, loads the holes in the way the formation will load them when the well is brought into production and while it produces.
6
A production well with a production casing is being abandoned. The formation at 1930–1950 m is perforated over 1934–1946 m. What plug do the rules of Kazakhstan require if the well is ordinary, and how does the requirement change if the reservoir fluid holds 5% hydrogen sulfide?
Show the answer
An ordinary well: the plug covers the perforated interval and another 20 m below and above it, from 1914 to 1966 m (52 m). It is checked by setting down the weight of the pipe. With hydrogen sulfide at 3.5% or more, the well is complex. It is abandoned under a project, and the productive formation is covered over its whole thickness and 100 m above its top, from 1950 to 1830 m (120 m). The wellhead is fitted with a casing head and a valve with outlets for monitoring the pressure. The rules require a corrosion-resistant plugging material in both cases, and with hydrogen sulfide it is especially important.

Updated: 19 October 2026