Paraffin & Wax Deposition

Overview

Crude oils contain dissolved paraffin waxes — long-chain n-alkanes typically in the C18 to C75+ range. As oil flows from the warm reservoir toward the surface, it cools. When the bulk oil temperature falls below the Wax Appearance Temperature (WAT, also called Cloud Point), wax crystals begin to nucleate and grow. These crystals deposit on cold pipe walls through molecular diffusion, forming an insulating gel layer that narrows the flow path, increases frictional pressure drop, and can ultimately plug the wellbore or flowline (Burger, Perkins & Striegler, 1981).

Wax deposition is distinct from gelation. Deposition is a gradual radial growth of solid wax on a cold surface driven by a temperature gradient. Gelation is the bulk solidification of the entire oil column when it cools below the pour point — a restart problem rather than a steady-state flow assurance problem. The two phenomena require different models and different treatments.

Theory

Wax Crystallization Thermodynamics

Wax precipitation is governed by solid-liquid equilibrium thermodynamics. Won (1986) modeled the wax phase as a non-ideal solid solution using modified regular solution theory, with vapor-liquid equilibria handled by the Soave-Redlich-Kwong equation of state. The key equilibrium condition for each component i is:

ln(x_i^S * gamma_i^S / x_i^L * gamma_i^L) = (delta_H_f,i / R) * (1/T_f,i - 1/T)

where x_i^S and x_i^L are solid and liquid mole fractions, gamma_i^S and gamma_i^L are activity coefficients, delta_H_f,i is the heat of fusion, T_f,i is the melting temperature of pure component i, R is the gas constant, and T is the system temperature. This framework predicts the WAT as the highest temperature at which any solid phase becomes thermodynamically stable.

Molecular Diffusion Mechanism

Burger, Perkins & Striegler (1981) identified three transport mechanisms for wax deposition: molecular diffusion, shear dispersion, and Brownian diffusion. Subsequent experimental work by Singh, Venkatesan & Fogler (2000, 2001) demonstrated that molecular diffusion is the dominant mechanism in most field conditions, while shear dispersion plays a secondary role.

The molecular diffusion mechanism works as follows: the radial temperature gradient in a pipe creates a radial concentration gradient of dissolved wax (because wax solubility decreases with temperature). Dissolved wax molecules diffuse toward the cold wall where they precipitate, building up the deposit layer.

Deposit Aging

Singh et al. (2000) showed that after initial deposition, the gel deposit ages through a counter-diffusion process. Higher carbon-number wax molecules diffuse into the deposit from the bulk oil, while lower carbon-number oil molecules diffuse out. This increases the wax fraction (F_w) and hardness of the deposit over time, making it more difficult to remove. A critical carbon number (CCN) exists: molecules above CCN diffuse in, molecules below CCN diffuse out.

Formulas

1. Wax Appearance Temperature (WAT) Estimation

No single empirical correlation reliably predicts WAT from API gravity alone — WAT depends strongly on the n-alkane distribution, which varies by crude source. However, a first-pass screening correlation based on wax content and API gravity (adapted from Weingarten & Euchner, 1988, and field data compilations) is:

WAT (deg F) approx = 80 + 1.2 * W_pct + 0.6 * (40 - API)

where:
  WAT      = Wax Appearance Temperature (deg F)
  W_pct    = wax content by weight (%)
  API      = API gravity of crude oil (dimensionless)

Warning: This screening correlation carries +/-15 deg F uncertainty. Laboratory DSC or cross-polarized microscopy measurement is required for design purposes. Do not use this for flowline insulation design without lab data.

2. Singh Molecular Diffusion Deposition Model

The Singh et al. (2000, 2001) model describes deposit thickness growth driven by radial mass transfer of dissolved wax molecules toward the cold pipe wall:

dh/dt = (1 / rho_dep) * D_wo * (dC_ws/dT) * (dT/dr)|_wall

where:
  dh/dt    = rate of deposit thickness growth (m/s)
  rho_dep  = density of the wax deposit (kg/m3)
  D_wo     = diffusion coefficient of wax in oil (m2/s)
  dC_ws/dT = temperature dependence of wax solubility (kg/m3/K)
  dT/dr    = radial temperature gradient at the wall (K/m)

The diffusion coefficient D_wo is typically calculated using the Wilke-Chang correlation (1955):

D_wo = 7.4e-8 * (phi_B * M_B)^0.5 * T / (mu_oil * V_A^0.6)

where:
  phi_B    = association parameter for the solvent (1.0 for hydrocarbons)
  M_B      = molecular weight of the solvent/oil (g/mol)
  T        = absolute temperature (K)
  mu_oil   = oil viscosity (cP)
  V_A      = molar volume of wax solute at normal boiling point (cm3/mol)

3. Deposit Aging (Wax Fraction Increase)

The wax fraction F_w in the deposit increases over time as higher carbon-number molecules counter-diffuse into the gel (Singh et al., 2001):

dF_w/dt = (D_eff / delta^2) * (C_ws(T_interface) - C_ws(T_wall))

where:
  F_w         = wax fraction in deposit (mass fraction, dimensionless)
  D_eff       = effective diffusivity of wax within the gel (m2/s)
  delta       = deposit thickness (m)
  C_ws(T)     = wax solubility concentration at temperature T (kg/m3)
  T_interface = temperature at bulk-oil / deposit interface (K)
  T_wall      = pipe wall temperature (K)

4. Matzain Semi-Empirical Model

The Matzain et al. (1999, 2001) model combines molecular diffusion with a shear-removal term. It is used in commercial simulators including OLGA Wax. The net deposition rate is:

dm_net/dt = C1 * dm_diffusion/dt  -  C2 * tau_w^C3

where:
  dm_net/dt       = net wax deposition mass rate (kg/m2/s)
  dm_diffusion/dt = molecular diffusion mass flux (kg/m2/s)
  tau_w           = wall shear stress (Pa)
  C1              = molecular diffusion tuning factor (default 15.0)
  C2              = shear removal coefficient (default 0.055)
  C3              = shear removal exponent (default 1.4)

The first term represents mass deposition by molecular diffusion (scaled by C1), and the second term represents mechanical removal by wall shear stress. At high flow rates, shear removal can exceed diffusion, leading to self-limiting deposit thickness.

5. Heat Transfer for Hot Oil Treatment

Hot oil (or hot water) treatment melts and re-dissolves deposited wax. The heat required to raise the deposit above its melting point is:

Q = m_dep * [Cp_wax * (T_melt - T_dep) + delta_H_f + Cp_oil * (T_treat - T_melt)]

where:
  Q         = total heat required (BTU or kJ)
  m_dep     = mass of wax deposit (lb or kg)
  Cp_wax    = specific heat of solid wax, typically 0.5 BTU/lb/F (2.1 kJ/kg/K)
  T_melt    = wax melting temperature (deg F or K)
  T_dep     = in-situ deposit temperature (deg F or K)
  delta_H_f = latent heat of fusion of wax, typically 80-100 BTU/lb (185-230 kJ/kg)
  Cp_oil    = specific heat of hot oil, typically 0.45-0.55 BTU/lb/F
  T_treat   = hot oil treatment temperature (deg F or K)

The hot oil circulation rate must supply heat faster than it is lost to the formation. The required flow rate:

q_hot = Q / (rho_oil * Cp_oil * (T_in - T_out) * t_treat)

where:
  q_hot   = hot oil circulation rate (bbl/hr or m3/s)
  T_in    = hot oil inlet temperature (deg F)
  T_out   = return temperature (deg F)
  t_treat = treatment duration (hr)

6. Pigging Frequency Optimization

The optimal pigging interval balances production losses (from flow restriction) against pigging costs. The maximum allowable deposit thickness before pigging is typically 10–25% of the pipe inner diameter. The time to reach that restriction:

t_pig = delta_max / (dh/dt)_avg

where:
  t_pig      = pigging interval (days)
  delta_max  = maximum allowable deposit thickness (inches or mm)
             = f_restrict * ID / 2
  f_restrict = fractional restriction limit (typically 0.10 to 0.25)
  ID         = pipe inner diameter (inches or mm)
  (dh/dt)_avg = average deposition rate (inches/day or mm/day)

Production loss due to wax restriction follows from the Hagen-Poiseuille relation — flow rate scales as (effective radius)^4:

q_restricted / q_clean = ((r - delta) / r)^4

where:
  r     = clean pipe inner radius
  delta = deposit thickness

7. Treatment Cost Optimization

The total cost of wax management per unit time is the sum of production losses and treatment costs. The objective is to find the treatment interval t* that minimizes total cost:

minimize  C_total(t) = C_prod_loss(t) + C_treatment / t

C_prod_loss(t) = integral from 0 to t of:
    P_oil * q_0 * [1 - ((r - dh/dt * tau) / r)^4] d(tau)

C_treatment = C_pig + C_chemical + C_rig_time

where:
  C_total      = total cost per unit time ($/day)
  P_oil        = oil price ($/bbl)
  q_0          = clean-well production rate (bbl/day)
  C_pig        = cost per pigging run ($)
  C_chemical   = chemical inhibitor cost per cycle ($)
  C_rig_time   = rig/equipment time cost per intervention ($)

Worked Example

Given: A producing well with API 28, 12% wax content, 8,000 ft depth, BHT = 190 deg F, WHT = 85 deg F, 2-7/8" tubing (ID = 2.441 in), producing 150 BOPD.

Step 1: Estimate WAT

WAT = 80 + 1.2 * 12 + 0.6 * (40 - 28)
    = 80 + 14.4 + 7.2
    = 101.6 deg F  (approx 102 deg F)

Check: For a 28 API crude with 12% wax, a WAT around 100–110 deg F is physically reasonable. Lab DSC data for similar crudes typically fall in the 95–115 deg F range.

Step 2: Determine Deposition Risk Zone

The temperature profile along the tubing (assuming linear gradient for screening):

dT/dz = (BHT - WHT) / depth = (190 - 85) / 8000 = 0.0131 deg F/ft

Temperature reaches WAT at depth:
z_WAT = (BHT - WAT) / (dT/dz) = (190 - 102) / 0.0131 = 6,718 ft from bottom

Deposition risk zone: from 6,718 ft to surface (top 1,282 ft of tubing)
  = upper 16% of the wellbore

Step 3: Estimate Deposition Rate

Using typical field values for 28 API crude:

D_wo   = 2.5e-10 m2/s   (Wilke-Chang at ~100 deg F for C25 in crude)
dC/dT  = 0.015 kg/m3/K  (typical for 12% wax crude near WAT)
dT/dr  = 500 K/m         (estimated from wellbore heat loss model)
rho_dep = 850 kg/m3      (wax-oil gel, ~60% wax fraction)

dh/dt = (1/850) * 2.5e-10 * 0.015 * 500
      = 2.21e-9 m/s
      = 0.19 mm/day
      = 0.0075 in/day

Step 4: Pigging Interval

Allowable restriction: 15% of ID
delta_max = 0.15 * 2.441 / 2 = 0.183 inches

t_pig = 0.183 / 0.0075 = 24.4 days

Production loss at pigging time:
q/q_0 = ((1.221 - 0.183) / 1.221)^4 = (0.850)^4 = 0.522

  --> 48% production loss at time of pigging (too high)

Step 5: Recommendation

Result: At 0.0075 in/day deposition rate in 2-7/8" tubing, production drops to ~52% of clean-well rate within 24 days. This well needs aggressive wax management:

• Pigging every 10–14 days to keep restriction below 8% (production loss <25%)
• Chemical inhibitor injection (pour point depressant / crystal modifier) to reduce deposition rate by 40–60%
• Consider downhole heater or insulated tubing if interventions exceed $3,000/month at 150 BOPD

Valid Ranges

Assumptions and Limitations

1. The WAT screening correlation is empirical and carries +/-15 deg F uncertainty. It is not a substitute for laboratory DSC or cross-polarized microscopy measurement.
2. The Singh diffusion model assumes a steady-state radial temperature gradient and single-phase oil flow. Multiphase flow (gas-liquid) alters both heat transfer and shear, requiring models such as Matzain or OLGA Wax.
3. Molecular diffusion dominates in laminar and moderate turbulent flow. At very high Reynolds numbers (Re > 50,000), shear removal may partially offset deposition (Matzain et al., 2001).
4. The Matzain tuning constants (C1=15, C2=0.055, C3=1.4) are defaults; they must be calibrated to field or loop data for each crude oil system.
5. Deposit aging is not modeled in steady-state screening. Fresh deposits (hours old) may contain only 10–20% wax and are soft; aged deposits (days to weeks) contain 50–80% wax and are hard.
6. The pigging frequency formula assumes uniform deposition along the risk zone. In practice, deposition is heaviest near the WAT isotherm and at low-velocity points (risers, bends).
7. Chemical inhibitors (PPDs, crystal modifiers) reduce deposition rate by 40–70% in lab tests, but field performance varies with inhibitor-crude compatibility.
8. All models require crude-specific wax solubility curves (C_ws vs. T) from laboratory measurements — these cannot be reliably predicted from bulk properties alone.

References

1. Burger, E.D., Perkins, T.K. & Striegler, J.H. (1981). “Studies of Wax Deposition in the Trans Alaska Pipeline.” Journal of Petroleum Technology, 33(6), 1075–1086. SPE-8788-PA.
2. Singh, P., Venkatesan, R., Fogler, H.S. & Nagarajan, N. (2000). “Formation and Aging of Incipient Thin Film Wax-Oil Gels.” AIChE Journal, 46(5), 1059–1074.
3. Singh, P., Venkatesan, R. & Fogler, H.S. (2001). “Morphological Evolution of Thick Wax Deposits during Aging.” AIChE Journal, 47(1), 6–18.
4. Won, K.W. (1986). “Thermodynamics for Solid Solution-Liquid-Vapor Equilibria: Wax Phase Formation from Heavy Hydrocarbon Mixtures.” Fluid Phase Equilibria, 30, 265–279.
5. Matzain, A., Apte, M.S., Zhang, H.-Q., Volk, M., Brill, J.P. & Creek, J.L. (2001). “Multiphase Flow Wax Deposition Modeling.” Presented at the ETCE 2001 Conference, ETCE2001-17114.
6. Weingarten, J.S. & Euchner, J.A. (1988). “Methods for Predicting Wax Precipitation and Deposition.” SPE Production Engineering, 3(1), 121–126. SPE-15654-PA.
7. Azevedo, L.F.A. & Teixeira, A.M. (2003). “A Critical Review of the Modeling of Wax Deposition Mechanisms.” Petroleum Science and Technology, 21(3–4), 393–408.
8. Theyab, M.A. (2018). “Wax Deposition Process: Mechanisms, Affecting Factors and Mitigation Methods.” Open Access Journal of Science, 2(2), 112–118.
9. Wilke, C.R. & Chang, P. (1955). “Correlation of Diffusion Coefficients in Dilute Solutions.” AIChE Journal, 1(2), 264–270.