Drilling guides › Annular velocity and equivalent circulating density (ECD)
Annular velocity and equivalent circulating density (ECD)
While the pumps run, friction in the annulus adds pressure on top of the hydrostatic. The well "feels" a heavier mud: the equivalent circulating density. Annular velocity tells you whether the cuttings are being lifted. The two pull in opposite directions: more flow cleans the hole better but raises the ECD, so a good flow rate is always a compromise between hole cleaning and the strength of the weakest formation.
Annular velocity
Or from pump output in bbl/min: AV = Pump output (bbl/min) ÷ Annular capacity (bbl/ft).
The two forms are the same sum. Dividing gallons per minute by 42 gives barrels per minute, and the annular capacity is (Hole ID² − Pipe OD²) ÷ 1029.4 bbl/ft. Flow divided by capacity is feet per minute, and 1029.4 ÷ 42 = 24.5. The result is an average: the mud moves faster in the middle of the gap and slower against the pipe and the wall.
Worked example
428 gpm around 5 in drill pipe in an 8½ in hole:
Check: 10.18 bbl/min ÷ 0.0459 bbl/ft ≈ 222 ft/min.
Typical targets in vertical hole are roughly 100–200 ft/min, depending on the mud and hole angle; high-angle wells usually need more flow and pipe rotation to clean the hole.
Worked example: one flow rate, three velocities
850 gpm in a 12¼ in hole with 5 in drill pipe and 8 in collars, drilled from a floating rig with a 19.5 in ID riser:
The slowest section controls hole cleaning. In this case cuttings that travel well up the open hole would settle in the riser, which is why riser boost pumps exist. The same logic applies to large casing above a smaller liner.
Annular velocity is only half of the cleaning story. Cuttings also slip back down through the mud at a rate that depends on their size and on the mud's viscosity and gel strength. Net lift is annular velocity minus slip velocity, which is why thin mud at a high rate and thicker mud at a lower rate can both clean a vertical hole. In deviated hole cuttings drop onto the low side and form beds, and only flow plus pipe rotation stirs them back into the stream.
Equivalent circulating density
Annular pressure loss (APL) comes from hydraulics software, your mud company's report, or the hydraulics screen of a drilling app. Only the annular part counts: the pressure lost inside the drill string and across the bit does not act on the open hole.
The formula is just the equivalent mud weight idea applied to circulation. Friction in the annulus has to be overcome to push mud up to the flowline, and that extra pressure is felt at every depth below. Expressing it as a mud weight lets you set it directly against the pore pressure and the fracture gradient.
Worked example
MW 10.5 ppg, TVD 9,500 ft, annular pressure loss 350 psi:
Worked example: ECD at the shoe
The casing shoe is at 5,000 ft TVD, the mud is 10.5 ppg and the hydraulics model gives 150 psi of annular loss between the shoe and surface.
Use only the friction above the point you are checking. The loss below the shoe raises the pressure at the bit, not at the shoe.
In metric: ECD (sg) = MW (sg) + APL (bar) ÷ (0.0981 × TVD (m)). For 1.25 sg mud, 3,000 m TVD and 25 bar of annular loss, ECD = 1.25 + 25 ÷ 294.3 ≈ 1.33 sg.
Why ECD matters
- Losses: if the ECD at the weakest point (often the casing shoe) goes above the fracture gradient or leak-off EMW, the well can start taking mud.
- Kicks: when the pumps stop, the bottomhole pressure falls back to hydrostatic. A well that is only stable while circulating can flow on connections.
- Tight margins: in deepwater and depleted zones the window between pore pressure and fracture pressure can be smaller than the ECD increase itself.
What raises the ECD
Higher flow rate, higher yield point and plastic viscosity, small clearances (large collars, liners, tight hole), a heavy load of cuttings in the annulus, and pipe rotation in narrow holes all add annular friction. Cuttings also raise the effective density of the returning mud, which adds to the hydrostatic part. Fast drilling with too little flow is a classic way to pack the annulus and break down the shoe.
Common mistakes
- Using the whole standpipe pressure, or the bit pressure drop, as if it acted on the formation.
- Dividing by measured depth. The friction is calculated along the measured length of the annulus, but the conversion to ppg uses TVD.
- Ignoring the riser, booster flow and surface choke when comparing simulated and measured pressures on floating rigs.
Questions people ask
Why does a downhole pressure tool read more than the hydraulics model?
The usual reasons are cuttings in the annulus that the model did not include, mud properties downhole that differ from the surface sample, and a hole that is smaller or more tortuous than assumed. A steady gap between the two is information, not noise.
Is there an ECD while tripping?
Yes, in effect. Running pipe in pushes mud up the annulus and adds friction (surge); pulling out drags mud down and takes pressure away (swab). Both behave like circulating at a rate set by the pipe speed, which is why trip speed limits are set from the same hydraulics.
Good practice
- Bring the pumps up in steps after connections and trips so that gelled mud does not give a pressure spike at the shoe.
- Compare the hydraulics model with a downhole pressure tool when one is in the string, and recalibrate the model when they disagree.
- Watch torque, drag and pressure trends for signs of cuttings beds, and circulate the hole clean before tripping.