Oilwell cements and additives

Quick refresher

The clinker of Portland cement is mainly four minerals. Alite ( $\mathrm{C_3S}$ ) gives early strength. Belite ( $\mathrm{C_2S}$ ) gives late strength. Aluminate ( $\mathrm{C_3A}$ ) reacts very fast; gypsum holds it back, and it decides the resistance to sulfates. Aluminoferrite ( $\mathrm{C_4AF}$ ) reacts moderately.
After mixing, the slurry hardly changes for several hours (the induction period). This is when the crew pumps and displaces it. Then hydration speeds up: the slurry thickens, sets and hardens. Temperature and pressure shorten this period.
The main well cements are classes G and H of API Spec 10A (in GOST 1581, types I-G and I-H). They have no additives, and the test water is 44 and 38% of the cement mass. The properties for a particular well are set with additives.
We calculate the density and the yield of a slurry from absolute volumes: the total mass of the components divided by the sum of their volumes. Neat class G cement (cement and water only) gives about 1.90 g/cm³ and 0.76 m³ of slurry per tonne.
Accelerators and retarders control the thickening time. Lightening and weighting additives control the density. Fluid-loss additives and dispersants control filtration and viscosity. Fillers help against lost circulation. Special systems are used against gas migration.
Above about 110 °C, set cement without silica loses strength and becomes permeable. 35–40% of silica by the mass of cement prevents this.

Key relationships

\rho_{\text{sl}} = \dfrac{\sum m_i}{\sum m_i/\rho_i}

density of a cement slurry from the absolute volumes of its components

Y = \dfrac{\sum m_i/\rho_i}{m_{\text{cem}}}

yield of the slurry per unit mass of cement: m³/t, with masses in tonnes and densities in t/m³

V_{\text{API}} = 2\,V_t\sqrt{30/t}

fluid loss index for 30 min, from the volume of filtrate $V_t$ collected in $t$ minutes in a cell of half the standard area

p = 0.00981\sum_j \rho_j h_j

pressure of the columns of fluids, MPa ( $\rho_j$ in g/cm³, $h_j$ are vertical heights in m)

Common pitfalls

Testing the thickening time at the wrong temperature. For pumping, what matters is the circulating temperature at the bottom (the temperature while fluid is pumped), which is lower than the static one (the undisturbed temperature of the rock). Strength is tested at the static temperature.
Lightening a slurry just with extra water. Without a filler the water separates, and the set cement is weak and permeable.
Forgetting that an additive does not act on one property only. Retarders thin the slurry, bentonite lowers strength, and sodium chloride speeds setting at low concentrations and slows it at high ones.
Going without silica in hot intervals. Strength falls not at once but after weeks and months, when it is hard to put right.
Assuming that fluid loss matters only for squeeze cementing (forcing cement into a leak under pressure).

After this lesson you can

explain which minerals make up the clinker of Portland cement (the burnt and ground raw material), and how hydration (the reaction with water) goes, which decides both the time when the slurry can be pumped and the strength of the set cement
tell the cement classes of API Spec 10A and the types of GOST 1581 apart, and read their designations
list the properties of the slurry and of the set cement (density, water–cement ratio and yield, thickening time, fluid loss, free fluid, strength) and explain how they are tested. The yield is the volume of slurry that a unit mass of cement gives. The thickening time is how long the slurry stays pumpable. Free fluid is water that separates from the slurry
calculate from a recipe the density and the yield of a slurry by the absolute volume method, and the consumption of materials for a given volume
choose families of additives for a task: accelerators (they speed up setting), retarders (they slow it down), lightening and weighting additives, fluid-loss additives, dispersants, fillers against lost circulation (mud escaping into the rock), and additives against gas migration (gas rising through the cement)
explain why above about 110 °C set cement loses strength, and how much silica is needed to prevent it

What a well needs from cement

The cement sheath (the ring of set cement) between the casing and the wall of the hole separates the formations. It does not let fluids flow between them or rise to the surface along the annulus. It holds the casing and protects it from outside against aggressive formation waters. For the conductor and the surface casing, it also holds the weight of the wellhead and blowout preventer equipment (strings by purpose). Cement is also needed for repair, to seal a leaking casing or an unwanted perforation (a hole made through the casing). It is needed for plugs too, when the crew tests a zone, sidetracks (drills a new branch from the well) or abandons the well.

The requirements for cement conflict with each other, and the cement must meet them in two stages:

While the slurry is liquid, it must stay pumpable throughout pumping and displacement, with a margin. It must have a density that holds the formations back but does not fracture weak ones. It must not give up its water to permeable rock, and it must not separate its own water. It must displace the drilling mud without mixing with it. It must not let gas through while it thickens.
When it has set, it must gain strength fast, so that the rig does not stand idle in waiting on cement (WOC). It must have low permeability. It must keep its strength and its tightness at formation temperature for decades. It must resist hydrogen sulfide, carbon dioxide and the sulfates of formation waters, and the changes of pressure and temperature in production.

A laboratory chooses the recipe from the data of the well: the temperature, the pressures, the density of the mud and the pumping time. It tests the same batch of cement, the same additives and the same mix water that will be delivered to the rig. The properties of different batches of one class differ noticeably, and fresh, formation or sea water affects setting in different ways.

Portland cement: clinker and hydration

Portland cement is made by burning a mixture of limestone and clay at a temperature of about 1450 °C. The partly melted mass fuses into granules of clinker, which are then ground with a small amount of gypsum. Almost all the clinker is made of four minerals. Cement chemistry writes them in a short form: C is CaO, S is SiO₂, A is Al₂O₃, F is Fe₂O₃ and H is H₂O.

Clinker minerals

Mineral: Alite, tricalcium silicate, $\mathrm{C_3S}$
Share in classes G and H: about a half
How it reacts with water: fast; gives the main strength in the first day

Mineral: Belite, dicalcium silicate, $\mathrm{C_2S}$
Share in classes G and H: from a quarter to a third
How it reacts with water: slowly; strength grows over weeks

Mineral: Tricalcium aluminate, $\mathrm{C_3A}$
Share in classes G and H: up to 8%, and up to 3% in high-sulfate-resistant cements
How it reacts with water: very fast and with a lot of heat; gypsum holds it back

Mineral: Calcium aluminoferrite, $\mathrm{C_4AF}$
Share in classes G and H: about a tenth
How it reacts with water: moderately; has little effect on strength

Mineral	Share in classes G and H	How it reacts with water
Alite, tricalcium silicate, $\mathrm{C_3S}$	about a half	fast; gives the main strength in the first day
Belite, dicalcium silicate, $\mathrm{C_2S}$	from a quarter to a third	slowly; strength grows over weeks
Tricalcium aluminate, $\mathrm{C_3A}$	up to 8%, and up to 3% in high-sulfate-resistant cements	very fast and with a lot of heat; gypsum holds it back
Calcium aluminoferrite, $\mathrm{C_4AF}$	about a tenth	moderately; has little effect on strength

The shares are approximate and differ between plants and batches. The limits on $\mathrm{C_3A}$ are set for the grades of moderate and high sulfate resistance.

Hydration of calcium silicates (a conventional notation)

2\,\mathrm{C_3S} + 6\,\mathrm{H} \rightarrow \mathrm{C_3S_2H_3} + 3\,\mathrm{CH}, \qquad 2\,\mathrm{C_2S} + 4\,\mathrm{H} \rightarrow \mathrm{C_3S_2H_3} + \mathrm{CH}

where:

$\mathrm{C_3S_2H_3}$ calcium silicate hydrate (C–S–H), a gel that fills the gaps and glues the grains together; its real composition varies, and the CaO/SiO₂ ratio in it is about 1.7, —
$\mathrm{CH}$ calcium hydroxide Ca(OH)₂, portlandite, —
$\mathrm{H}$ water, —

Units: molar amounts in the notation of cement chemistry

Holds when

Conventional stoichiometry: C–S–H has no constant composition.
For temperatures up to about 110 °C. At higher temperatures other products form (see the section on high temperatures).

The silicates turn into C–S–H gel, which gives strength and tightness (no flow through the cement), and into crystals of portlandite. Portlandite makes the pore water (the water in the pores) strongly alkaline (pH about 13). This protects the steel of the casing from corrosion, but portlandite is also the first to react with carbon dioxide and sulfates.

Aluminate $\mathrm{C_3A}$ without gypsum sets almost at once. Gypsum (calcium sulfate) binds it into needle-like crystals of ettringite on the surface of the grains, and slows the reaction. The sulfates of formation water cause the same reaction in the set cement. The growing crystals push it apart and destroy it. So against sulfate waters we need a cement with a low content of $\mathrm{C_3A}$ .

SchematicThe rate of heat release of a hardening cement slurry against time (a diagram): a short burst after mixing, the induction period, when the slurry is pumped, the speed-up of hydration with thickening and setting, then a slowing down and a long hardening.

If we measure the heat release of a hardening slurry, we see several periods. Right after mixing there is a short burst: the grains get wet, and the first reactions happen. Then comes the induction period: there is almost no heat, the slurry stays liquid, and this is when the crew pumps and displaces it. Then the hydration of the silicates speeds up sharply: the slurry thickens, sets and starts to gain strength. After a day or two the rate falls, but hardening goes on for weeks. A rise in temperature and pressure shortens the induction period. Retarders lengthen it, and accelerators shorten it.

The reactions themselves need little water, about a quarter of the mass of the cement. We take more (for classes G and H, 44 and 38%), so that the slurry can be mixed and pumped. The water above what is needed stays in the set cement as pores: the more of it there is, the weaker and more permeable the set cement is.

Classes and types of oilwell cements

A well cement differs from a building cement first of all in being the same from batch to batch: each batch must give a predictable thickening time and strength. The requirements for it are set by the standard API Spec 10A (ISO 10426-1), which divides cements into classes.

Cement classes of API Spec 10A

Class: A
Use: ordinary conditions with no special requirements; historically down to a depth of about 1830 m (6000 ft)
Water in tests, % of the cement mass: 46

Class: B
Use: the same, with moderate or high sulfate resistance
Water in tests, % of the cement mass: 46

Class: C
Use: high early strength: fine grinding, more $\mathrm{C_3S}$
Water in tests, % of the cement mass: 56

Class: D
Use: moderately high temperature and pressure, historically about 1830–3050 m; retarders are allowed at the factory
Water in tests, % of the cement mass: 38

Class: G and H
Use: the main well cements: no additives except gypsum and water; by themselves down to a depth of about 2440 m (8000 ft), and with additives in almost any conditions
Water in tests, % of the cement mass: 44 (G), 38 (H)

Class	Use	Water in tests, % of the cement mass
A	ordinary conditions with no special requirements; historically down to a depth of about 1830 m (6000 ft)	46
B	the same, with moderate or high sulfate resistance	46
C	high early strength: fine grinding, more $\mathrm{C_3S}$	56
D	moderately high temperature and pressure, historically about 1830–3050 m; retarders are allowed at the factory	38
G and H	the main well cements: no additives except gypsum and water; by themselves down to a depth of about 2440 m (8000 ft), and with additives in almost any conditions	44 (G), 38 (H)

From API Spec 10A, 25th edition (2019). Sulfate resistance is marked O (ordinary), MSR (moderate) and HSR (high). Classes E and F for deep hot wells were in earlier editions of the standard and are not in the one in force.

Classes G and H are close in chemical composition. H is ground more coarsely, so it needs less water, and the slurry is heavier: about 1.98 g/cm³ against 1.90 for G. Class G is the most common well cement in the world: additives set its properties for almost any conditions.

In the post-Soviet countries, oilwell Portland cements are made to the interstate standard GOST 1581. Kazakhstan was among the countries that adopted its 1996 edition. In Russia the 2019 edition has been in force since 2022. The standard distinguishes:

Type I: Portland cement with no additives. I-G and I-H are cements with no additives, tested at a water–cement ratio of 0.44 and 0.38. They are the counterparts of classes G and H.
Type II: with mineral additives (6–20%). Type III: with special additives, including lightweight cements (marked Ob, a transliteration of the Russian abbreviation; slurry density from 1.10 to 1.60 g/cm³) and weighted cements (marked Ut, from 2.00 to 2.30 g/cm³).
The temperature of use for types I–III: (50) is 15 to 50 °C, (100) is 51 to 100 °C, and (150) is 101 to 150 °C.
Sulfate resistance: SS (transliterated from the Russian letters). For types I-G and I-H, SS-1 (high) and SS-2 (moderate).

A designation reads from left to right. PCT-I-G-SS-1 is oilwell Portland cement of type I-G with high sulfate resistance. PCT-III-Ob5-100 is a lightweight one, with a slurry density of about 1.50 g/cm³, for moderate temperatures. (The letters are transliterated from the Russian: PCT stands for the Russian words for “oilwell Portland cement”.) The project for building the well states which standard and type is used on a particular well.

For special conditions there are other binders. High-alumina cement, based on calcium aluminates, is used in permafrost (ground that stays frozen for years), where Portland cement sets too slowly, and in wells with a very high temperature. There are also mixtures of Portland cement with pozzolans (materials that react with the lime of the cement) and slags, and polymer compositions for repair work.

Density, water–cement ratio and yield

We set the amount of water by the water–cement ratio (W/C): the mass of water per unit mass of cement. A ratio of 0.44 means 44 kg of water per 100 kg of cement. Dry additives are dosed as a percentage of the cement mass (BWOC, by weight of cement), salts often as a percentage of the water mass (BWOW), and liquid additives in litres per tonne of cement (in English-language practice, in gallons per sack).

We calculate the density and the volume of a slurry by the absolute volume method. Each component takes in the slurry the volume of its own substance, which is its mass divided by its true density. It does not take the bulk volume of the powder, which holds a lot of air. The volume of the slurry is the sum of these volumes, and its density is the total mass divided by this volume.

Density of a cement slurry from absolute volumes

\rho_{\text{sl}} = \frac{\sum_i m_i}{\sum_i m_i/\rho_i}

where:

$\rho_{\text{sl}}$ density of the cement slurry, g/cm³
$m_i$ mass of a component: cement, water, each additive, kg
$\rho_i$ true (absolute) density of a component: cement about 3.15, water 1.00, silica flour about 2.65, g/cm³ (= kg/L)

Units: kg and kg/L: the volumes come out in litres, and the density in kg/L = g/cm³

Holds when

The volumes of the components add up. Dissolving and chemical shrinkage on mixing are not included.
Additives that dissolve in water are counted by their own density. The error is small, because the amounts are small.
There is no entrained air in the slurry. In foaming, the density is lower than the calculated one, so the crew checks it on the rig with a pressurised mud balance.

Components that are lighter than the slurry (water, fillers, microspheres) lower the density, and heavier ones (cement, weighting agents) raise it. Small additives of a fraction of a per cent hardly change either the density or the volume.

In other unit systems (1)

Oilfield units

\rho\,[\text{ppg}] = \frac{\sum m_i\,[\text{lb}]}{\sum m_i/(8.345\,\rho_i)\;[\text{gal}]}

masses in pounds per sack of cement (94 lb, 42.64 kg), $\rho_i$ is the specific gravity, and the volume is in US gallons.

8.345 is the number of pounds in a gallon of a substance with a density of 1 g/cm³; 1 g/cm³ = 8.345 ppg.

Slurry yield and consumption of materials

Y = \frac{\sum_i m_i/\rho_i}{m_{\text{cem}}}, \qquad M_{\text{cem}} = \frac{V}{Y}

where:

$Y$ slurry yield, the volume of slurry per unit mass of cement, m³/t
$m_{\text{cem}}$ mass of cement in the recipe, t
$V$ the required volume of slurry (from the cementing design), m³
$M_{\text{cem}}$ mass of cement for the job; we find the water and the additives by multiplying by their share of the cement mass, t

Units: masses in tonnes, densities in t/m³ (numerically equal to g/cm³)

Holds when

The same as for the density. A margin for losses on mixing and for the inaccuracy of the hole volume is added separately.

The yield shows how much slurry a tonne of cement gives. Fillers and water increase the yield, and weighting agents decrease it. So lightweight slurries are also cheaper per cubic metre.

In other unit systems (1)

Oilfield units

Y\,[\text{ft}^3/\text{sack}] = \frac{\sum m_i/(8.345\,\rho_i)\;[\text{gal}]}{7.48}

masses per sack of cement of 94 lb; 7.48 gallons in a cubic foot.

Conversion: 1 m³/t = 1.506 ft³ per sack.

CalculatedThe calculated density of a cement slurry against the water–cement ratio: neat cement (3.15 g/cm³) and the tail slurry recipe of the example with 35% silica flour. The dots are the standard water for classes H and G and the recipe of the example.

Why not just add water to lighten a slurry? Cement holds only a limited amount of water. If there is more, the particles settle, water separates on top, and the set cement is weak and permeable. So we lighten a slurry with fillers that bind water or are light themselves (the section on additives). In the other direction, if we reduce the water and add a dispersant, we can make a slurry heavier without a weighting agent, up to about 2.16 g/cm³ (18 ppg). Beyond that a weighting agent is needed.

Worked exampleTail slurry (the last batch of cement pumped) for a hot interval: density, yield, materials

Given (assumed data)

Class G cement: 100 kg (3.15 g/cm³)
Silica flour (finely ground quartz): 35% of the cement mass (2.65 g/cm³)
Fluid-loss additive and retarder: 0.6% (1.30 g/cm³) and 0.3% (1.50 g/cm³) of the cement mass
Mix water: 53 kg per 100 kg of cement (W/C 0.53), from the laboratory design
Slurry volume needed: 15 m³ (from the cementing design)

Solution

Absolute volumes per 100 kg of cement: cement $100/3.15 = 31.75$ L, silica flour $35/2.65 = 13.21$ L, fluid-loss additive $0.6/1.30 = 0.46$ L, retarder $0.3/1.50 = 0.20$ L, water 53 L. The sum is 98.6 L.
Mass $100 + 35 + 0.6 + 0.3 + 53 = 188.9$ kg. Density $188.9/98.6 = 1.916$ g/cm³, or $1.916\cdot 8.345 = 16.0$ ppg.
Yield: 98.6 L per 100 kg of cement, that is 0.986 m³/t (1.48 ft³ per sack). The mix water is 0.53 m³/t.
For 15 m³ of slurry: cement $15/0.986 = 15.21$ t, silica flour $0.35\cdot 15.21 = 5.32$ t, water $0.53\cdot15.21 = 8.06$ m³, fluid-loss additive 91 kg, retarder 46 kg.
For comparison, neat class G cement at W/C 0.44: $\dfrac{144}{31.75 + 44} = 1.901$ g/cm³, and the yield is 0.76 m³/t.

Answer: The slurry has a density of 1.916 g/cm³ and a yield of 0.986 m³/t. For 15 m³ we need 15.21 t of cement, 5.32 t of silica flour and 8.06 m³ of water.

What it means. Silica flour needs its own water, so the W/C is higher than for neat cement. But quartz is dense (2.65 g/cm³), and the density of the slurry has hardly changed, while the volume of slurry per tonne of cement has grown by 30%: the same tonne of cement carries more solids and water. Additives of a fraction of a per cent hardly changed the density and the yield, because their volume is less than 1% of the slurry. But they set the thickening time and the fluid loss. Materials are taken to the rig with a margin for losses on mixing and for the inaccuracy of the hole volume. The densities of the materials are typical, and the recipe is assumed for illustration.

Tests: thickening time, fluid loss, free fluid, strength

We test a slurry so that the test repeats what the slurry will “see” in the well: the pressure and the temperature during pumping (the methods are standardised in API RP 10B-2, ISO 10426-2). We find the thickening time at the bottomhole circulating temperature (BHCT). While the slurry is pumped, circulation has cooled the hole, and in deep wells this temperature is noticeably lower than the static one. We find the strength at the bottomhole static temperature (BHST), because the set cement hardens when the well warms up again.

Main tests of a cement slurry

Property: Thickening time
How it is measured: a consistometer (a device that stirs the slurry under well pressure and temperature and measures how thick it is); the time to a consistency of 100 Bc
Guide value: the time of the job with a margin set by the programme

Property: Fluid loss
How it is measured: filtration through a 45 µm screen at a pressure difference of 6.9 MPa for 30 min
Guide value: without a fluid-loss additive, hundreds of mL or more; where there is a danger of gas flow, 50 mL or less

Property: Free fluid
How it is measured: the slurry stands for 2 h in a measuring cylinder, sometimes tilted at 45°
Guide value: for inclined and gas wells, practically zero

Property: Compressive strength
How it is measured: cubes with a 50.8 mm edge crushed in a press, or an ultrasonic analyser, which does not destroy the sample
Guide value: at the time of drilling out, often not less than 3.5 MPa (500 psi)

Property: Rheology (how the slurry flows)
How it is measured: a rotational viscometer
Guide value: for calculating pressures and the pumping regime

Property	How it is measured	Guide value
Thickening time	a consistometer (a device that stirs the slurry under well pressure and temperature and measures how thick it is); the time to a consistency of 100 Bc	the time of the job with a margin set by the programme
Fluid loss	filtration through a 45 µm screen at a pressure difference of 6.9 MPa for 30 min	without a fluid-loss additive, hundreds of mL or more; where there is a danger of gas flow, 50 mL or less
Free fluid	the slurry stands for 2 h in a measuring cylinder, sometimes tilted at 45°	for inclined and gas wells, practically zero
Compressive strength	cubes with a 50.8 mm edge crushed in a press, or an ultrasonic analyser, which does not destroy the sample	at the time of drilling out, often not less than 3.5 MPa (500 psi)
Rheology (how the slurry flows)	a rotational viscometer	for calculating pressures and the pumping regime

Bc (Bearden units of consistency) are conventional units of consistency from 0 to 100. 70 Bc is also often taken as the limit of pumpability. Practice in the CIS countries also measures the spread of the slurry on a cone and the setting times.

Thickening time is the time in which the slurry at well conditions becomes unpumpable. It must be longer than the time of the job (mixing, pumping, displacement), with a margin for stops. If the slurry thickens sooner, it stays in the casing, which is one of the worst accidents in cementing. A time that is too long is also harmful: the wait on cement (WOC) is longer, the slurry stays liquid for longer opposite gas-bearing formations, and solid particles have time to settle. It is best when the slurry stays liquid for a long time and then thickens fast. This curve is called “right-angle set”.

SchematicThe consistency of a slurry in a consistometer test (a diagram): ordinary gradual thickening and “right-angle” thickening. The thickening time is the time to 100 Bc; it must cover the time of the job with a margin.

Fluid loss is the volume of filtrate (the liquid that passes through) that the slurry gives up through a filter under a pressure difference. Against a permeable formation, the slurry loses water and thickens, and it may turn into a dense plug. The annulus will not fill completely, the casing may stick, and in narrow gaps around liners the slurry will not pass at all. The test is like the high-pressure, high-temperature test of drilling mud (filtration test). The area of the filter is half the standard one, so the volume is doubled.

Fluid loss index of a cement slurry

V_{\text{API}} = 2\,V_t\,\sqrt{\frac{30}{t}}

where:

$V_{\text{API}}$ fluid loss index: the volume of filtrate for 30 min, reduced to the standard area, mL/30 min
$V_t$ the volume of filtrate collected in the test by the time $t$ , mL
$t$ the time of the test: 30 min, or the moment when the sample dehydrated and gas started to go through the filter, min
$2$ the conversion from the area of the cell (about 22.6 cm²) to the standard one, twice as large, —

Units: mL and minutes

Holds when

The filtrate grows in proportion to $\sqrt t$ , as in filtration through a growing cake (filtration theory).
When the sample dehydrates in minutes, this is only a conventional index for comparing recipes.

The smaller the index, the longer the slurry keeps its water against a permeable formation. The doubling and the square root of time are a conversion to the standard area and time. They are not physical properties of the slurry.

Worked exampleFluid loss of two samples: without a fluid-loss additive and with one

Given (assumed data)

Sample 1 (no fluid-loss additive): dehydrated (lost all its free water) after 1.5 min; 74 mL collected
Sample 2 (with 0.6% additive): 23 mL collected in 30 min
Limit of the programme: 50 mL/30 min

Solution

Sample 1: $V_{\text{API}} = 2\cdot 74\cdot\sqrt{30/1.5} = 2\cdot 74\cdot 4.472 = 662$ mL/30 min.
Sample 2: $V_{\text{API}} = 2\cdot 23 = 46$ mL/30 min.

Answer: Without the additive the index is about 662 mL/30 min, and with it 46 mL/30 min, which is within the limit of the programme.

What it means. Sample 1 dehydrated in a minute and a half. Against a permeable sandstone such a slurry would quickly turn into a dense mass. The additive lowered the fluid loss more than 14 times. For an ordinary string a few hundred millilitres would be enough. But for a liner with a narrow gap, or where gas may flow, we need a fluid loss of about 50 mL/30 min or less. The data in this example are assumed for illustration.

Free fluid is the water that collects above a slurry at rest. In a vertical well it rises to the top of the cement column. In an inclined well it collects along the upper wall and leaves a through channel along the whole length of the interval. So for inclined and gas wells we make the slurry stable: with no water separation and no settling of particles.

Strength is needed to hold the string and to withstand the blows of the bit when it drills out the float collar and the shoe. The set cement must also not break up during perforation (shooting holes through the casing and the cement) and hydraulic fracturing (cracking the formation with pumped fluid). How fast it grows depends on temperature. At the bottom of a hot well the slurry gains 3.5 MPa in a few hours, and at the wellhead of a cold surface casing it takes much longer. This is what decides the wait on cement (WOC). An ultrasonic analyser lets us follow the gain of strength continuously, without destroying the sample.

Additives: what each family controls

Almost every cement slurry contains additives. Dry ones are mixed with the cement at the depot, and liquid ones and salts are usually added to the mix water. Their action depends on temperature, on the batch of cement and on the other additives, so the recipe is tested as a whole.

Accelerators: They shorten the induction period and speed up the gain of strength at low temperatures: in shallow strings, in winter, in permafrost. The most common is calcium chloride (usually 2–4% of the cement mass). Sodium chloride speeds setting at low concentrations and slows it at high ones (above about 18–20% of the water mass). Sodium silicate and sea water are also used.
Retarders: They lengthen the induction period at high temperatures. They are lignosulfonates, hydroxycarboxylic acids and their salts, cellulose derivatives, synthetic polymers and phosphonates, and at very high temperatures these with intensifiers. The action depends sharply on temperature: a dose calculated for the bottom will delay the setting of the top of the cement column, which is colder, for a long time. Sugar is a very strong retarder, and accidental contamination with it is dangerous.
Extenders (lightening additives, fillers): They lower the density and increase the yield. Some bind extra water: bentonite (a clay that swells in water), sodium silicate. Others are light themselves: pozzolans (fly ash, diatomite), hollow glass or aluminosilicate microspheres. For foamed cements, nitrogen is added. With bentonite or silicate we get slurries down to about 1.4–1.5 g/cm³, and with microspheres to about 1.0–1.2 g/cm³ (glass ones are lighter than aluminosilicate ones). Foamed slurries are lighter still. The price is a lower strength that comes more slowly, and with microspheres, the crushing of some particles under high pressure.
Weighting agents: They raise the density against high-pressure formations: hematite (4.9–5.1 g/cm³; usually up to 2.3 g/cm³ of slurry, and higher with a dispersant), ilmenite (about 4.5), manganese tetroxide (about 4.8: very fine, so it settles less). Barite (4.2–4.3) needs a lot of water and lowers strength, so it is rarely used in cements.
Fluid-loss additives: Polymers: cellulose derivatives, hydroxyethyl cellulose (HEC) and its carboxymethyl derivative (CMHEC), synthetic copolymers based on AMPS (2-acrylamido-2-methylpropane sulfonic acid), polyvinyl alcohol, latex. Bentonite lowers fluid loss only a little.
Dispersants (plasticisers): Polynaphthalene sulfonates, lignosulfonates and polycarboxylates. They give the particles the same electric charge, so the particles repel each other. Viscosity and yield point (the force needed to start the slurry flowing) fall, and the slurry goes into turbulent flow at a lower pressure and displaces the drilling mud better. We can use less water and make the slurry heavier without a weighting agent. An overdose causes settling of particles and water separation.
Lost-circulation materials: Granular (gilsonite, nut shells, crushed coal), flaky (cellophane, mica) and fibrous (synthetic fibres) fillers plug cracks and channels. Thixotropic slurries (they turn into a gel when they stand still) thicken fast at rest, so they do not flow away into the formation.
Gas-migration control: Latex, microsilica and other fine particles that lower the permeability of a thickening slurry; expanding additives; recipes with a short change from liquid to solid; very low fluid loss and zero free fluid (the section on gas migration).
Others: Defoamers work against foaming when mixing. Silica works against the loss of strength at high temperatures (the next section). Reinforcing fibres and elastic additives serve set cement that goes through cycles of pressure and temperature. Salt is used for cementing in salt formations.

Worked exampleA lightweight lead slurry (the first batch pumped) and a weak formation

Given (assumed data)

Lead slurry: class G cement + water 75 kg per 100 kg of cement + hollow microspheres (0.60 g/cm³); the density needed is 1.45 g/cm³
Well: shoe at 2400 m; the tail slurry of the example above (1.916 g/cm³) from 2100 to 2400 m, the lead slurry from 1200 to 2100 m, and above it drilling mud of 1.22 g/cm³
Weak formation: 2150 m, fracture pressure (the pressure at which the rock breaks) equal to a mud density of 1.50 g/cm³ (an equivalent mud weight)

Solution

The mass of microspheres $m$ per 100 kg of cement, from the condition $\dfrac{100 + 75 + m}{31.75 + 75 + m/0.60} = 1.45$ : $m = \dfrac{175 - 1.45\cdot 106.75}{1.45/0.60 - 1} = 14.27$ kg, that is 14.27% of the cement mass.
Volume $106.75 + 14.27/0.60 = 130.5$ L per 100 kg of cement. The yield is 1.305 m³/t, which is 1.32 times more than for the tail slurry (0.986 m³/t).
If the whole annulus were filled with the tail slurry, the pressure on the weak formation after placement would be $0.00981\,(1.22\cdot1200 + 1.916\cdot950) = 32.2$ MPa, and the equivalent mud weight would be 1.53 g/cm³, which is more than the fracture pressure.
With the lightweight lead slurry: $0.00981\,(1.22\cdot1200 + 1.45\cdot900 + 1.916\cdot50) = 28.1$ MPa, and the equivalent mud weight is 1.33 g/cm³. The margin to fracture is 0.17 g/cm³.

Answer: The lead slurry with 14.27% of microspheres has a density of 1.45 g/cm³ and a yield of 1.305 m³/t. With it, the pressure on the weak formation after placement corresponds to 1.33 g/cm³ instead of 1.53 g/cm³.

What it means. This is a static check after the end of displacement. During pumping, the friction pressure losses in the annulus are added to it (Circulating pressure losses), and the margin of 0.17 g/cm³ is needed precisely for them. The check during the job is made in the cementing design (Primary cementing). The heavy strong slurry is left at the shoe, where the set cement carries the load while the next interval is drilled, and the light lead slurry fills the top of the annulus. Under pressure some microspheres may crush, and the density of the slurry in the well will be higher than the calculated one, so their strength is chosen with a margin. The data in this example are assumed for illustration. The lesson Fracture gradient and leak-off testing explains how the fracture pressure is found.

High temperature: loss of strength and silica

Up to about 110 °C the C–S–H gel is stable. Above it, with no additives, the gel slowly turns into dense crystals (it recrystallises) of α-dicalcium silicate hydrate ( $\alpha\text{-}\mathrm{C_2SH}$ , with a CaO/SiO₂ ratio of 2). The crystals take less volume than the gel, and the bonds between them are weaker. Over weeks and months the set cement loses a large part of its strength, and its permeability grows by orders of magnitude (many times over). This is strength retrogression, a silent threat. In a test at 1–3 days the recipe looks good, but half a year later the cement sheath no longer isolates the formations.

The remedy has been known for a long time: add 35–40% of silica by the cement mass, as silica flour (finely ground quartz) or silica sand. The extra silica reacts with the calcium and lowers the CaO/SiO₂ ratio to about one. Then on heating, other crystalline silicate hydrates form: tobermorite, and at higher temperatures xonotlite. They keep their strength and their low permeability.

What forms in set cement on heating

Temperature: up to ~110 °C
Without silica: C–S–H gel and portlandite: the set cement is strong
With 35–40% silica: the same; the silica hardly reacts

Temperature: ~110–150 °C
Without silica: α-C₂SH: strength falls, permeability grows
With 35–40% silica: tobermorite (CaO/SiO₂ ≈ 0.83): strength is kept

Temperature: above ~150 °C
Without silica: α-C₂SH and other phases: strength and impermeability are lost even faster
With 35–40% silica: xonotlite (CaO/SiO₂ = 1) and related phases: the set cement is stable

Temperature	Without silica	With 35–40% silica
up to ~110 °C	C–S–H gel and portlandite: the set cement is strong	the same; the silica hardly reacts
~110–150 °C	α-C₂SH: strength falls, permeability grows	tobermorite (CaO/SiO₂ ≈ 0.83): strength is kept
above ~150 °C	α-C₂SH and other phases: strength and impermeability are lost even faster	xonotlite (CaO/SiO₂ = 1) and related phases: the set cement is stable

The temperature limits are approximate: the transformations take time, and their speed depends on the pressure, the composition of the cement and the fineness of the silica. The calculation of where 35–40% comes from is in the advanced section.

We add silica when the static temperature in the cemented interval is above about 110 °C, or will rise above it later: with steam injection, or in geothermal wells. Fine flour reacts faster. Coarse sand reacts more slowly and affects the properties of the slurry less. For very high temperatures the fineness of the silica, and the mix of flour and sand, are chosen by tests. At the same time, such recipes need strong retarders that keep working at these temperatures, and long-term tests at formation conditions.

Contamination and gas migration

Contamination by drilling mud. Drilling mud and cement slurry are often incompatible: their mixture thickens sharply, and the chemicals of the mud, such as thinners, can slow down setting. Contaminated cement hardens badly and does not bond to the casing and the rock. So a spacer (a fluid that separates the two) and wiper plugs (plugs that wipe the inside of the pipe and keep the fluids apart) keep the cement apart from the drilling mud, and the laboratory checks the mixtures for compatibility (Primary cementing).

Gas migration. While the slurry is liquid, it passes the full hydrostatic pressure (the pressure of its own column) to the formations. When it starts to thicken, its gel strength grows: the structure of the slurry leans on the walls and the casing, and it stops passing the weight of the column down. At the same time the slurry loses volume: filtrate leaves, and chemical shrinkage goes on. The pressure against a gas-bearing formation falls, and if it falls below the formation pressure, gas enters the slurry that has not yet set and makes channels in it. Then the gas rises along them to the wellhead. This is how pressures in the annuli between strings, and craters, appear.

The dangerous time is when the slurry no longer passes pressure, but is not yet impermeable. We estimate it by the transition time: from a gel strength of about 48 Pa (100 lbf/100 ft²), when the transmission of pressure is noticeably disturbed, to about 239 Pa (500 lbf/100 ft²), when gas can no longer pass. The shorter this time, the lower the risk. So against gas-bearing formations we use recipes with a short transition time, very low fluid loss and no free fluid, and also latex, expanding and other impermeable systems.

Terms and symbols

Cement slurry and set cement: The mixture of cement, water and additives that is pumped into the well (cement slurry), and what it hardens into (set cement).
Mix water: The water (fresh, sea, or with dissolved additives) used to prepare the cement slurry.
Water–cement ratio (W/C): The mass of water per unit mass of cement. For classes G and H in tests it is 0.44 and 0.38.
Slurry yield: The volume of slurry per tonne of cement (in English-language practice, per sack of 94 lb).
Thickening time: The time in which the slurry at well conditions becomes unpumpable (to 100 Bc in a consistometer).
Fluid loss: The volume of filtrate that the slurry gives up under a pressure difference. For cement it is measured at 6.9 MPa for 30 min.
Free fluid: The water that separates above a slurry at rest.
WOC: Waiting on cement: the time while the set cement gains the strength needed to continue the work.
BHCT and BHST: The bottomhole circulating and static temperatures. The first is used to test the thickening time, and the second to test the strength.
Strength retrogression: The loss of strength and the growth of permeability of set cement above about 110 °C when there is no silica.
Transition time: The time when a thickening slurry no longer passes the full pressure, but does not yet stop gas from passing.

Common mistakes and misconceptions

Wrong

“Thickening time is the time in which the cement sets.”

In fact

It is the time while the slurry can be pumped. The set cement gains the strength needed to continue the work later, from a few hours to a day or more, depending on temperature.

Wrong

“More water is easier to pump, so it is better.”

In fact

The extra water separates, the particles settle, and the set cement is weak and permeable. We lighten a slurry with fillers and improve pumpability with dispersants.

Wrong

“Fluid loss matters only for squeeze cementing.”

In fact

In primary cementing too, a slurry that gives up water against permeable formations thickens, plugs narrow gaps and makes gas migration easier. For liners and gas wells, we lower the fluid loss especially strongly.

Wrong

“A retarder is enough for a hot well.”

In fact

A retarder gives the time for pumping, but it does not save the set cement from strength retrogression above about 110 °C. The cement also needs 35–40% of silica by the cement mass.

Wrong

“Cement slurry is weighted with barite, like drilling mud.”

In fact

Barite needs a lot of water and lowers the strength of the set cement. For cement we use hematite, ilmenite or manganese tetroxide, and up to about 2.16 g/cm³ we manage by reducing the water with a dispersant.

In depth: advanced detail

Why exactly 35–40% silica

The dangerous phase $\alpha\text{-}\mathrm{C_2SH}$ forms when there is a lot of calcium in the silicates: the CaO/SiO₂ ratio is close to 2. The stable tobermorite ( $\mathrm{C_5S_6H_5}$ , ratio 0.83) and xonotlite ( $\mathrm{C_6S_6H}$ , ratio 1.0) need about one mole of SiO₂ for each mole of CaO. Let us calculate how much silica to add, from the composition of the silicates of the clinker.

Worked exampleCaO/SiO₂ ratio of the silicates of cement with quartz

Given (assumed data)

Cement: 55% $\mathrm{C_3S}$ and 25% $\mathrm{C_2S}$ by mass (typical shares, assumed)
Molar masses: CaO 56.08, SiO₂ 60.08, $\mathrm{C_3S}$ 228.32, $\mathrm{C_2S}$ 172.24 g/mol
Quartz: 35 or 40 g per 100 g of cement

Solution

In 55 g of $\mathrm{C_3S}$ : CaO $55\cdot168.23/228.32 = 40.52$ g, SiO₂ $55\cdot60.08/228.32 = 14.47$ g. In 25 g of $\mathrm{C_2S}$ : CaO $25\cdot112.15/172.24 = 16.28$ g, SiO₂ 8.72 g.
In total 56.80 g of CaO (1.0128 mol) and 23.19 g of SiO₂ (0.3860 mol): the ratio is 2.62.
With 35 g of quartz: $1.0128/((23.19 + 35)/60.08) = 1.05$ . With 40 g it is 0.96.

Answer: Silica lowers the CaO/SiO₂ ratio in the silicates from 2.62 to 1.05 at 35% and to 0.96 at 40%, that is, to about one, as in tobermorite and xonotlite.

What it means. This is where the rule of 35–40% comes from. With less quartz there is an excess of calcium, and some α-C₂SH is left. With much more, the extra quartz only dilutes the set cement. The calculation assumes that all the quartz has reacted, and that the calcium of the aluminates and ferrites does not take part in these reactions. With silicate shares of 50 and 30% instead of 55 and 25%, the result is almost the same. Fine flour reacts more fully, and coarse sand more slowly, so for very high temperatures sometimes more is added. The composition of the cement is assumed for illustration.

Connections to other topics

Casing: purpose, types, properties and standards — the strings that are cemented, the shoe, the float valve and the centralisers.
Primary cementing — placing the slurry behind the casing, volumes, displacement, spacers.
Remedial cementing, cement plugs and cement evaluation — squeeze cementing, plugs and the evaluation of the quality of cementing.
Fracture gradient and leak-off testing — the fracture pressure is the upper limit of the density of a cement slurry.
Pore pressure and overburden pressure — hydrostatic pressure and equivalent mud weight.
Drilling fluids: functions and properties — the density, weighting and filtration of drilling mud: compare with cement.
Field testing of drilling fluids — the filter press and the theory of filtration through a cake.
Drilling fluid rheology and flow regimes — the rheological models that also describe a cement slurry.
Well control: primary barrier and kick indicators — gas migration and complications of cementing as a threat to well control.

Check yourself

1
Which clinker mineral gives the main strength of set cement in the first day?
1. A.Alite, $\mathrm{C_3S}$
2. B.Belite, $\mathrm{C_2S}$
3. C.Tricalcium aluminate, $\mathrm{C_3A}$
4. D.Aluminoferrite, $\mathrm{C_4AF}$
Show the answer
Correct answer: A. Alite, $\mathrm{C_3S}$
Alite reacts with water fast and gives the C–S–H gel, which gains strength in the first day. Belite reacts slowly and adds strength over weeks. Aluminate reacts very fast, but it adds little to strength, and gypsum holds it back.
2
Find the density and the yield of a neat slurry of class H cement at the standard water–cement ratio of 0.38 (cement density 3.15 g/cm³), and compare it with class G.
Show the answer
The volume per 100 kg of cement is $100/3.15 + 38 = 31.75 + 38 = 69.75$ L, and the mass is 138 kg. The density is $138/69.75 \approx 1.98$ g/cm³ (about 16.5 ppg), and the yield is 0.70 m³/t. For class G at 0.44 it is 1.90 g/cm³ and 0.76 m³/t. Class H is heavier and gives less slurry per tonne, because it needs less water.
3
At what temperature is the thickening time of a slurry for cementing a string tested?
1. A.At the temperature at the wellhead
2. B.At the static formation temperature at the bottom
3. C.At the circulating temperature at the bottom
4. D.At room temperature
Show the answer
Correct answer: C. At the circulating temperature at the bottom
While the slurry is pumped, circulation has cooled the hole, and the slurry is at the circulating temperature, which is lower than the static one. A test at the static temperature would give too short a time and too much retarder. A test at room temperature would give a dangerously overestimated time. Strength, on the other hand, is tested at the static temperature.
4
In a 30-minute fluid loss test, 31 mL of filtrate was collected. What is the index, and does the slurry meet the programme limit of 50 mL/30 min?
Show the answer
The cell has half the standard area, so the index is $2\cdot 31 = 62$ mL/30 min. This is more than 50. We need more fluid-loss additive, or another additive, and then the test is repeated.
5
Why do we add 35–40% of silica flour by the cement mass to the recipe for an interval with a static temperature of 140 °C?
1. A.So that on heating, tobermorite and xonotlite form instead of α-C₂SH, and the set cement does not lose strength
2. B.To lengthen the thickening time
3. C.To lower the density of the slurry
4. D.To reduce the fluid loss
Show the answer
Correct answer: A. So that on heating, tobermorite and xonotlite form instead of α-C₂SH, and the set cement does not lose strength
Above about 110 °C, C–S–H gel without silica turns into α-C₂SH, and the set cement loses strength and impermeability. Silica lowers the CaO/SiO₂ ratio to about one, and stable tobermorite and xonotlite form. The thickening time is set with a retarder, and the fluid loss with a fluid-loss additive.
6
A cement slurry of 1.90 g/cm³ fills the annulus for 1500 m above a shoe at a depth of 2500 m, and above it there is drilling mud of 1.20 g/cm³. Will the formation at the shoe, with a fracture pressure of 1.55 g/cm³ as an equivalent mud weight, hold the static pressure after placement?
Show the answer
The equivalent mud weight at the shoe is $(1.90\cdot1500 + 1.20\cdot1000)/2500 = 4050/2500 = 1.62$ g/cm³. This is more than 1.55, so the formation will fracture, the slurry will be lost, and the cement will not rise to the height needed. The top of the column is made lightweight, or the height is reduced, or the job is done in two stages. During pumping, friction is added to the static pressure.

Updated: 15 October 2026