Cement placement: ECD, free-fall and the lift pressure

A primary cement job is over in two hours and cannot be done again. In that time a train of fluids passes every weak zone in the open hole, the cement in the casing falls ahead of the pumps, and the final pressures tell you whether the cement went where it was meant to. This article works through a production casing job in a deep, hot tight-gas well, and a design that had to change to keep a weak aquifer intact.

What happens while the cement is placed

The job pumps spacer, lead and tail slurry down the casing, drops the top plug and displaces it with mud until it lands on the float collar. Each depth in the annulus feels the hydrostatic of whatever stands above it plus the friction of the returns flowing past. Expressed as a density, that is the job's ECD. It climbs as the heavy slurries rise up the annulus and peaks near the end of displacement, with the full cement column in place and the pumps still at rate.

Two effects set cementing apart from drilling. For much of the job the fluids inside the casing outweigh the mud outside, so casing and annulus act as a U-tube. When the inside column outweighs the annulus plus its friction, the fluid in the casing falls faster than the pumps feed it: surface pressure drops to zero, returns outrun the pumps and a gap opens at the top of the casing. That is free-fall. The pumps then have to fill the gap, and while they do, returns run below the pump rate.

At the end, the annulus is full of cement and the casing full of mud. The pumps must hold the difference, the lift pressure, plus the friction at the final rate: the final circulating pressure. The plug bumps a set margin above it. A final circulating pressure well below plan means less cement in the annulus than planned.

The method

DrillIQ's placement simulation moves the fluid train through the well step by step by pumped volume. Each fluid loses pressure by its own Herschel–Bulkley rheology, fitted to its viscometer readings, in the casing and in each annular section. At each step:

ρ_i, h_i

density (SG) and vertical height (m) of each fluid above depth z

ΔP_f,ann(z)

annular friction above z (bar); 0.0981 bar per metre per SG

P_h

hydrostatic pressure of the column (bar)

Q

pump rate; the bump margin here is 500 psi, 34.5 bar

In free-fall the simulation finds the return rate from the balance with zero pressure at the top of the falling column, tracks the gap and lets the pumps catch it up. It checks the highest ECD each depth sees against the fracture gradient and the lowest against the pore pressure, the spacer's contact time across each zone, and the gas flow potential and post-gel hydrostatic at gas zones. Fluids are incompressible, with no temperature effect on rheology.

The circulating temperature for the slurry tests comes from the Kutasov–Targhi correlation, following the approach of API RP 10B-2: BHCT = −50.64 + 804.9 Γ + (1.342 − 12.22 Γ) BHST, with Γ the geothermal gradient in °C/m.

The example well

The example is a vertical tight-gas well in conditions typical of the Cooper Basin, Australia. The gas sits in deep, low-permeability sandstones with mild overpressure. Above them lie the sandstone aquifers of the Great Artesian Basin, which the production casing cement must isolate. The rock is hot: the geothermal gradient is 4.0 °C per 100 m from a 27 °C surface. The 5½″ production casing is cemented from 3,600 m to 1,000 m, 200 m inside the 9⅝″ shoe, to carry the fracture stimulation that follows.

At 171 °C the cement needs silica. Portland cement loses strength and gains permeability above about 110 °C unless 35–40 % silica flour by weight of cement is added, which is why both slurries carry it.

A hand check

1. Capacities. π/4 × (8.5² − 5.5²) × 0.0254² = 0.02128 m³/m in open hole; 0.02422 m³/m inside the 9⅝″ (8.835″ ID); 0.01105 m³/m inside the casing (4.670″ ID).
2. Volumes. Lead: 200 m × 0.02422 + 1,700 m × 0.02128 × 1.2 = 48.3 m³. Tail: 700 m × 0.02128 × 1.2 plus the 24 m shoe track, 0.27 m³, = 18.1 m³. Displacement: 3,576 m × 0.01105 = 39.5 m³. The engine lays the same volumes and puts the top of cement at 1,000 m.
3. Static column at the aquifer. With 16 m³ of spacer standing 660 m above the cement, the column at 1,800 m is (340 × 1.20 + 660 × 1.35 + 800 × 1.45) / 1,800 = 1.366 SG. The engine's static column there at the end of the job is 1.366 SG.
4. The static differential. At the shoe the annulus weighs 527.9 bar and the casing, mud above a cement-filled shoe track, 425.3 bar: 102.6 bar of differential. The simulation's lift pressure, with the fluids laid exactly as pumped, is 101.6 bar, within 1 %.
5. Temperatures. BHST = 27 + 4.0 × 36 = 171 °C. With Γ = 0.04 °C/m, BHCT = −50.64 + 804.9 × 0.04 + (1.342 − 12.22 × 0.04) × 171 = 127.5 °C, the engine's figure.

The hand figures fix the static picture. They can't say how far above 1.366 SG the aquifer goes while the cement moves, or when.

The first plan and the aquifer

The team's rule is to keep the job's ECD at least 0.05 SG below the fracture gradient at every open-hole depth. For the aquifer that means 1.43 SG. The first plan, a 1.50 SG lead, 12 m³ of spacer and displacement at 1,800 L/min, puts the aquifer at 1.440 SG at the end of displacement: 0.040 SG under its fracture gradient, which breaks the rule. The same run flags a second problem: the spacer passes the gas sandstones for only 8.0 minutes, under the usual 10-minute guideline for mud removal.

Slowing the displacement or lightening the lead each clears the margin. More spacer fixes the contact time but costs margin, because 1.35 SG spacer replaces 1.20 SG mud above the cement: 20 m³ breaks the rule again. The final plan, a 1.45 SG lead, 16 m³ of spacer and displacement at 1,500 L/min, keeps 0.051 SG of margin and 10.8 minutes of contact for 7.4 minutes more job time.

For the first hour the aquifer sits at about 1.26 SG, the 1.20 SG mud above it plus its friction, while the free-falling cement is still inside the casing. It drops towards 1.20 SG while the pumps stop for the plug and the returns slow. It climbs only when the spacer, then the lead, come up the annulus past 1,800 m, and peaks in the last minutes at full displacement rate, when the whole cement column is in the annulus and moving. The final plan's peak, 1.429 SG, comes about seven minutes later than the first plan's because it pumps more spacer and displaces more slowly; slowing to 500 L/min for the last 3 m³ brings it down before the bump.

Over the whole open hole, the aquifer is the only tight point. The peak ECD at the 5½″ shoe is 1.566 SG against 1.85, and at the 9⅝″ shoe 1.371 SG against its 1.65 SG FIT. The notch in the fracture gradient at 1,750–1,800 m decides the design.

Free-fall, lift and the bump

Pressure at the cement head starts at 36.5 bar, the friction of circulating mud and spacer. As the 1.45 SG lead fills the casing it falls, and at 29.9 m³ pumped, 22.6 minutes in, it reaches zero: the cement is falling faster than the pumps feed it. Returns rise to 102.7 m³/h against 90 m³/h pumped, and two minutes into the 15-minute plug drop, with the pumps stopped, they are still coming at about 21 m³/h. At most 8.5 m³ of the casing's top stands empty. The column comes back into balance at 82.4 m³, and displacement at 1,500 L/min spends until 105.9 m³ filling the gap, with returns well below the pump rate and the gauge still at zero.

That stretch is where jobs get misread. Returns above the pump rate are not a kick, and returns below it during the catch-up are not losses. A crew that slows down or pumps extra displacement to make up "lost" volume changes the job. The plan tells them in advance: the gauge reads zero from 29.9 to 105.9 m³.

From 105.9 m³ the pressure climbs fast, to 135.7 bar at 118.5 m³, as the gap closes and the heavy annulus starts pushing back. At the final 30 m³/h the engine puts the final circulating pressure at 132.7 bar: 101.6 bar of lift and 31 bar of friction. The plug lands at 121.9 m³ and is bumped at 167.2 bar, 34.5 bar above that. The first plan, with its heavier lead, would have bumped at 174.0 bar.

Gas, temperature and time

After placement the cement gels and, before it sets, loses its ability to transmit full hydrostatic. The engine's post-gel check weighs the gelled cement as water. On that basis the column at the gas sandstones falls to 1.095 SG against 1.118 SG pore pressure: an underbalance in the worst case. The gas flow potential at the same zone, the ratio of the pressure the gel can hold back to the overbalance, is 2.72, in the "minor" class (up to 3). Read together, the risk is real but modest. Gas control then rests on the tail itself: a short transition from fluid to solid and low fluid loss.

Time is comfortable. At the 127.5 °C circulating temperature the tail is in the well for 63.5 minutes of the job, and needs 123.5 minutes with a 60-minute safety margin against its 4-hour lab thickening time. The lead's 95.7 minutes of exposure leave even more room against its 5½ hours.

What to do on the job

Plot the job live against the simulation. Pressure against volume pumped, beside the planned line, is the best single view of a cement job. Pressure that comes back from free-fall early suggests less cement in the casing than planned; a final circulating pressure well below 132.7 bar suggests losses, most likely at the aquifer.

Count volume. The catch-up is 23.5 m³, from 82.4 to 105.9 m³. Check pump strokes against a tank on the 39.5 m³ of displacement.

Slow down before the bump. The last 3 m³ at 500 L/min lowers the final ECD and lands the plug softly. If it doesn't bump, over-displace by no more than half the shoe track, 0.13 m³ here.

Read the final pressures. Final circulating pressure less friction is the lift, a measure of the cement column. Record both before you bump. For the drilling side of the same window, see the hydraulics article and the mud-weight window article; for the casing loads after cementing, the casing design article.

A checklist for the rig team

• Run the placement simulation with the real fluid rheologies and rates, not a friction allowance.
• Know the weakest open-hole depth, its fracture gradient and the job's peak ECD there.
• Tell the crew when the gauge will read zero and when returns will run above and below the pump rate.
• Check spacer contact time across every zone to be isolated.
• Test slurries at the circulating temperature the well will see, and add silica above about 110 °C.
• Plot measured pressure against the planned line through the whole job.
• Record the final circulating pressure and the bump, and compare the lift with plan.
• Check the post-gel column and the gas flow potential at every gas zone.

How DrillIQ handles it

In DrillIQ the Cementing module designs each job from the well's geometry, fluids and pump schedule: volumes, sacks and mix water per slurry, top of cement, temperatures, thickening-time checks and centralizer standoff. Its placement simulation gives the surface pressure, returns, free-fall, lift, final circulating and bump pressure, the ECD envelope against the window, spacer contact time, the post-gel check and the gas flow potential. On the day it plots the cement unit's live data against the plan; afterwards it reads the bond log and records the acceptance. Everything above was worked out with the same engine.

Book a demo → See all capabilities

References

1. American Petroleum Institute. Recommended Practice 10B-2 (ISO 10426-2): Recommended Practice for Testing Well Cements.
2. American Petroleum Institute. Specification 10A (ISO 10426-1): Cements and Materials for Well Cementing.
3. American Petroleum Institute. Recommended Practice 65-2: Isolating Potential Flow Zones During Well Construction.
4. Nelson, E.B. and Guillot, D. (eds.) Well Cementing, 2nd edition, 2006.
5. Bourgoyne, A.T., Millheim, K.K., Chenevert, M.E. and Young, F.S. Applied Drilling Engineering. SPE Textbook Series Vol. 2, 1986.

About this example. The well in this article is a representative scenario set in typical Cooper Basin, Australia conditions and worked through with DrillIQ's engines. It is not a client's well, and its figures are not field results.