By Saad Iqbal
Every CT job, someone on the crew flips through a laminated hydraulics chart, finds the row closest to the pump rate you’re actually running, and hopes the fluid in the chart is close enough to the 2% KCl brine in your tanks. Then the treating pressure comes back ten percent off from what the chart predicted, and nobody can say whether that’s the friction pressure being wrong, a restriction downhole, or just a bad chart. A coiled tubing friction pressure calculation you can trust — and rerun in seconds for any pump rate, fluid, or CT size — fixes that guessing game for good.
This tutorial walks through the governing equation, works a real CT job by hand so you can see every number, and then turns the whole thing into a Python function you can drop into your next job design spreadsheet or app.
What You’ll Need
- CT string data: outer diameter (OD) and nominal wall thickness, in inches
- Fluid data: density in pounds per gallon (ppg) and viscosity in centipoise (cp) — use the Newtonian approximation for water/brine; gelled or slickwater fluids need a power-law model, noted below
- Job data: planned pump rate in barrels per minute (bbl/min) and the measured depth (string length) in feet
- Python 3 (python.org) with no extra libraries required — or run it in Jupyter if you want an interactive notebook
The worked example below uses 1.75 in CT with a 0.087 in wall, 8.5 ppg 2% KCl brine (viscosity ~1.0 cp), a pump rate of 1.2 bbl/min, and a measured depth of 12,000 ft — a fairly typical CT clean-out job.
Step 1 — Convert Your Pump Rate to Fluid Velocity
Friction pressure depends on how fast the fluid is actually moving inside the pipe, not on the pump rate itself. First get the CT inner diameter from the OD and wall thickness:
d = OD − 2 × wall
d = 1.75 − 2(0.087) = 1.576 in
Then convert pump rate to velocity with the standard oilfield formula (q in bbl/min, d in inches, v in ft/s):
v = 17.15 × q / d²
v = 17.15 × 1.2 / 1.576² = 17.15 × 1.2 / 2.484 = 8.29 ft/s

Step 2 — Calculate the Reynolds Number
The Reynolds number tells you whether the flow is laminar or turbulent, which decides which friction-factor equation applies. In oilfield units (ρ in ppg, v in ft/s, d in inches, μ in cp):
NRe = 928 × ρ × v × d / μ
NRe = 928 × 8.5 × 8.29 × 1.576 / 1.0 = 103,004
Anything above roughly NRe = 4,000 is turbulent (below NRe = 2,100 is laminar, with a transition zone between). At 103,004 this job is comfortably turbulent — the normal regime for water-based CT circulating fluids.

Step 3 — Find the Friction Factor and Pressure Gradient
For turbulent flow in smooth pipe (Re between roughly 4,000 and 100,000), the Blasius correlation gives the Fanning friction factor directly — no iteration needed:
f = 0.0791 / NRe^0.25
f = 0.0791 / 103,004^0.25 = 0.004415
Now plug f into the Darcy-Weisbach friction pressure gradient, again in field units:
dP/dL (psi/ft) = f × ρ × v² / (25.8 × d)
dP/dL = 0.004415 × 8.5 × 8.29² / (25.8 × 1.576) = 0.0634 psi/ft
Multiply by the 12,000 ft measured depth for the total friction pressure loss through the string:
ΔPf = 0.0634 × 12,000 = 761 psi

That 761 psi is friction alone — add it to hydrostatic and any surface/backpressure terms to get your total treating pressure prediction.
Step 4 — Automate It in Python
Hand-calculating this for every rate change during a job isn’t realistic. Here’s a small Python function that reproduces the exact numbers above and handles both laminar and turbulent regimes automatically:
def ct_friction_pressure(od_in, wall_in, rate_bpm, depth_ft, density_ppg, visc_cp): """ Coiled tubing friction pressure via Darcy-Weisbach. od_in, wall_in : CT outer diameter and wall thickness (in) rate_bpm : pump rate (bbl/min) depth_ft : measured depth / string length (ft) density_ppg : fluid density (ppg) visc_cp : fluid viscosity (cp), Newtonian approximation Returns dict with velocity, Reynolds number, friction factor, gradient (psi/ft) and total friction pressure (psi). """ d_in = od_in - 2 * wall_in # CT ID, in v_fts = 17.15 * rate_bpm / d_in**2 # velocity, ft/s n_re = 928 * density_ppg * v_fts * d_in / visc_cp if n_re < 2100: f = 64 / n_re # laminar (Fanning) else: f = 0.0791 / n_re**0.25 # Blasius (turbulent) grad_psi_ft = f * density_ppg * v_fts**2 / (25.8 * d_in) dp_total = grad_psi_ft * depth_ft return { "id_in": round(d_in, 3), "velocity_fts": round(v_fts, 2), "reynolds": round(n_re, 0), "friction_factor": round(f, 5), "gradient_psi_ft": round(grad_psi_ft, 4), "friction_pressure_psi": round(dp_total, 1), }result = ct_friction_pressure( od_in=1.75, wall_in=0.087, rate_bpm=1.2, depth_ft=12000, density_ppg=8.5, visc_cp=1.0,)print(result)# {'id_in': 1.576, 'velocity_fts': 8.29, 'reynolds': 103004.0,# 'friction_factor': 0.00442, 'gradient_psi_ft': 0.0634,# 'friction_pressure_psi': 760.4}
Wrap the function in a loop over a range of depths and you get the full pressure-vs-depth profile for the job, which is exactly what the chart below shows:

If you’re running friction pressure calculations across many wells or job types as part of a bigger workflow, petropt is a petroleum-engineering Python library that bundles hydraulics, PVT and multiphase-flow functions like this one, so you don’t have to maintain your own copy of every correlation.
How Do You Know the Answer Is Right?
Sanity-check any friction pressure calculation against these three things before you trust it on location:
- Order of magnitude: for typical CT clean-out rates (1–2 bbl/min) and depths (8,000–15,000 ft), friction pressure usually lands somewhere between 400 and 1,500 psi with water-based fluids. If your number is 10x that, check your unit conversions first — psi/ft versus psi/1,000 ft is the most common mistake.
- Reynolds number regime: confirm you’re using the right friction-factor equation for the regime you actually landed in (laminar below 2,100, turbulent above roughly 4,000). The Python function above switches automatically, but a hand calc won’t catch a wrong assumption for you.
- Compare to a treating pressure trend: if you have offset job data at a similar rate and fluid, your calculated friction pressure plus hydrostatic should track the observed surface treating pressure within about 10–15%. A bigger gap usually points to a restriction, a different fluid rheology than assumed, or reeled-CT curvature effects (see below).
Common Pitfalls
- Reeled CT isn’t straight pipe. The Darcy-Weisbach model above assumes straight pipe flow. Coiling the string onto a reel and running it through a gooseneck adds extra friction in those sections — CT service company hydraulics charts typically apply a correction factor (often 1.1–1.3x) for the reeled and curved portions. This tutorial’s number is a solid baseline; treat it as a lower bound for real-world CT strings, not the final word.
- Non-Newtonian fluids need a different model. Slickwater, gelled fluids, and friction reducers don’t follow the simple viscosity term used here — you need a power-law or Herschel-Bulkley rheology model instead. Water and brine (as in this example) are close enough to Newtonian for the approximation to hold.
- Wrong constant for the units you’re in. The 17.15 and 25.8 constants only work with the exact unit set used here (bbl/min, inches, ft/s, ppg, cp). Swap in SI units and you need different constants entirely — don’t mix and match.
Once friction pressure is nailed down, the natural next step is checking whether your pump rate is actually cleaning the hole — see the related tutorial below on annular velocity and hole cleaning calculation, or work through casing burst pressure design if you’re checking the string on the other side of the job.
When Should You Use Haaland Instead of Blasius?
The worked example above uses the Blasius correlation for smooth-pipe turbulent flow and reports the Fanning friction factor. For rougher tubing or a wider Reynolds-number range, use the Haaland approximation for the Darcy friction factor:
1 / sqrt(f_D) = -1.8 log10( [((epsilon / D) / 3.7)^1.11] + 6.9 / Re )
Darcy and Fanning factors are not interchangeable: f_D = 4 f_F. Match the friction-factor convention to the pressure-drop equation. Mixing a Darcy factor with a Fanning-form equation overpredicts pressure loss by roughly fourfold. Also treat the straight-pipe result as a baseline because reel curvature and coiled-tubing geometry add secondary-flow losses.
