Gas condensate and wet gas reservoirs

The Caribou Ridge wells produce about 250 m³ of condensate per million m³ of gas (45 barrels of condensate per million standard cubic feet) in their first months. At current prices the condensate earns more per well than the gas does. Mei's laboratory has now finished the full PVT study on the A-12-3 recombined sample: a constant composition expansion, which found a dew point of 27,600 kPa (4,000 psia), and a constant-volume depletion test. The Montney at Caribou Ridge is a gas condensate, about 3,400 kPa (500 psi) above its dew point. Farid calls a meeting with Lena, Ethan and Mei. “Every well we drill will drop that liquid into the rock as soon as the pressure near the fractures falls below the dew point,” he says. “We need to know how much we lose and what it does to the gas rate.”

Wet gas: recombination and gas equivalent

For a wet gas, and for a gas condensate above its dew point, the reservoir contains a single gas phase. At the surface it splits into separator gas and stock-tank liquid. Every reservoir calculation must use the well stream, which includes the liquid's molecules in the gas phase. The well-stream gravity follows from a molar balance of the two surface streams, using the producing gas-liquid ratio R, the liquid specific gravity γ_o and its molar mass M_o. In field units, with R in scf/STB:

The term 132,800 γ_o/M_o is the gas equivalent of one stock-tank barrel of liquid: the standard cubic feet of gas it would form if vaporized. When M_o has not been measured, Cragoe's correlation, M_o = 6084/(API − 5.9), is a reasonable estimate for condensates. Gas in place from volumetrics or material balance is then a well-stream volume, which must be split into dry gas and condensate in place using the gas equivalent.

# Wet gas recombination (separator gas gravity behind a known well-stream gravity) and condensate recovery from a constant-volume depletion (CVD) test
# of the liquids-rich Caribou Ridge / Ten Mile Ridge gas condensate.
SI = True      # True: metric units, False: oilfield units

gw, api, cgr0 = 0.74, 52.0, 45.0               # recombined well-stream gravity (Lesson 1), condensate API, initial CGR (bbl/MMscf)
go = 141.5 / (131.5 + api)                      # condensate specific gravity
Mo = 6084 / (api - 5.9)                         # Cragoe's molar mass correlation
R = 1e6 / cgr0                                  # producing gas-condensate ratio, scf/STB
geq = 132800 * go / Mo                          # gas equivalent of one barrel of condensate, scf
gg_sep = (gw * (R + geq) - 4584 * go) / R      # separator gas gravity that recombines to gw
print(f"Condensate gravity {go:.3f}, molar mass {Mo:.0f} g/mol")
print(f"Separator gas gravity {gg_sep:.3f} for a well-stream gravity of {gw:.2f}; one {'m3' if SI else 'bbl'} of condensate = "
      + (f"{geq * 0.0283168 / 0.158987:,.0f} m3 of gas" if SI else f"{geq:,.0f} scf of gas"))
print(f"Condensate share of well-stream moles: {geq / (R + geq):.1%}")
# CVD report: pressure psia, liquid dropout % of cell volume, cumulative well-stream produced (% of initial moles),
# and condensate-gas ratio of the gas removed at that step (bbl/MMscf)
cvd = [(4500, 0.0, 0.0, 45.0), (4000, 0.0, 4.2, 45.0), (3500, 4.5, 14.6, 33.0), (3000, 9.8, 26.8, 23.0),
       (2500, 13.2, 39.0, 16.5), (2000, 14.6, 51.2, 12.5), (1500, 14.1, 63.3, 10.2), (1000, 12.8, 75.4, 9.0),
       (700, 11.9, 82.5, 8.8)]
G = 60.7e9                                      # initial well-stream gas in one spacing unit, scf (Lesson 3)
cond0 = G * (1 - geq / (R + geq)) / 1e6 * cgr0  # initial condensate in place, STB
cum_c = 0.0; prev = cvd[0]
print(" p kPa   dropout %  well stream %  condensate recovered %" if SI else
      " p psia  dropout %  well stream %  condensate recovered %")
for row in cvd:
    p, sl, np_, cgr = row
    dG = (np_ - prev[2]) / 100 * G * (1 - geq / (R + geq))
    cum_c += dG / 1e6 * (cgr + prev[3]) / 2
    prev = row
    pp = f"{p / 0.145038:6,.0f}" if SI else f"{p:6,.0f}"
    print(f"{pp}  {sl:8.1f}  {np_:11.1f}  {100 * cum_c / cond0:12.1f}")
k = 0.158987 if SI else 1.0
u = "10^3 m3" if SI else "MSTB"
print(f"Condensate in place {cond0 * k / 1000:,.0f} {u}; recovered by depletion to {cvd[-1][0] / 0.145038 if SI else cvd[-1][0]:,.0f} "
      f"{'kPa' if SI else 'psia'}: {cum_c * k / 1000:,.0f} {u} ({cum_c / cond0:.0%})")

Condensate gravity 0.771, molar mass 132 g/mol
Separator gas gravity 0.607 for a well-stream gravity of 0.74; one m3 of condensate = 138 m3 of gas
Condensate share of well-stream moles: 3.4%
 p kPa   dropout %  well stream %  condensate recovered %
31,026       0.0          0.0           0.0
27,579       0.0          4.2           4.2
24,132       4.5         14.6          13.2
20,684       9.8         26.8          20.8
17,237      13.2         39.0          26.2
13,789      14.6         51.2          30.1
10,342      14.1         63.3          33.1
 6,895      12.8         75.4          35.7
 4,826      11.9         82.5          37.1
Condensate in place 420 10^3 m3; recovered by depletion to 4,826 kPa: 156 10^3 m3 (37%)

The Caribou Ridge condensate has an API gravity of 52 and a molar mass of about 132 g/mol, and each cubic metre of condensate is equivalent to about 138 m³ of gas. Although it makes up only 3.4 percent of the molecules, the condensate carries a large share of the value. The well-stream gravity of 0.74, from Lesson 1, corresponds to a separator gas of gravity about 0.61. An engineer who computed the reservoir z-factor from 0.61 instead of 0.74 would overstate it by about 3 percent and miss the liquid entirely.

Retrograde condensation

Ordinary condensation happens when a gas is compressed or cooled. Retrograde condensation is the opposite: liquid forms as pressure falls at constant temperature. It occurs because the reservoir temperature lies between the mixture's critical temperature and its cricondentherm. Below the dew point, the heavy components no longer fit in the gas phase as the gas expands and their solvent, the dense high-pressure methane, becomes less dense. As pressure falls further, the liquid volume reaches a maximum and then, at low pressure, part of it revaporizes, because light components boil off from the liquid and the heavier ones become more volatile relative to the thin gas.

Two laboratory tests describe this. The constant composition expansion holds the sample in a cell and lowers pressure in steps without removing anything; it measures the dew point and the relative volume. The constant-volume depletion, CVD, mimics a depleting reservoir: after each pressure step, gas is removed from the cell until its volume returns to the original value, as if produced, while the liquid stays behind, as it would in the rock. The CVD measures the liquid dropout, the two-phase z-factor, the composition of each removed gas and the cumulative production at each step. It is the standard test for gas condensate reservoir engineering.

The CVD of the A-12-3 sample shows liquid dropout rising to about 15 percent of pore volume at 13,800 kPa (2,000 psia) and only slight revaporization below that. Very little of this liquid will ever flow: the critical condensate saturation, below which condensate is immobile, is typically 10 to 30 percent in conventional rock and higher in tight rock. The condensate-gas ratio of the produced gas falls as the heavy ends stay behind, from 250 at the dew point to under 60 m³ per million m³ (10 bbl/MMscf) at low pressure. Integrating that falling ratio over the produced gas, as the code does, shows that depletion to 4,800 kPa (700 psia) recovers only about 37 percent of the original condensate, while it recovers more than 80 percent of the gas.

Condensate banking and well productivity

The liquid that drops out does not form evenly through the reservoir. Near a producing well, pressure falls below the dew point long before the average reservoir pressure does. A ring of condensate builds up around the wellbore or along the hydraulic fractures, where the gas flowing through deposits liquid continuously until the saturation is high enough for the condensate to flow. Inside this condensate bank the gas relative permeability can fall to 0.1 to 0.3 of its single-phase value. Fevang and Whitson (1996) described three regions: the near-well bank, where both phases flow; an intermediate region where condensate accumulates but does not flow; and the outer region still above the dew point.

Condensate banking can cut a gas well's productivity by half or more in moderate-permeability rock, and the effect in tight rock with hydraulic fractures is a subject of active study. Two effects partly offset it. At high velocity, the capillary number near the well is large enough to reduce the condensate saturation and raise the gas relative permeability, a positive effect called velocity stripping. And the inertial non-Darcy pressure loss competes with the bank for the same small region. Simulation of the near-well region needs fine grids or special well models, and relative permeability data measured at realistic capillary numbers.

Managing a gas condensate reservoir

The classic answer to retrograde loss is gas cycling: produce the well stream, strip out the condensate and NGLs in a plant, and inject the dry residue gas back into the reservoir to keep its pressure above the dew point. Cycling recovered large volumes of liquids from conventional condensate fields in Alberta, Texas, the North Sea and elsewhere in the decades when gas markets were limited. Its cost is the deferral of gas sales for many years, so it pays only when liquids are valuable, gas is cheap or unmarketable for a time, and the reservoir is permeable enough for injected gas to sweep it. Partial pressure maintenance, lean gas injection after depletion and nitrogen injection are variations.

Cycling is not practical in the Montney, where permeability is in the microdarcies and the injected gas would barely move between wells. The Caribou Ridge team's levers are different. The first is completion design: longer laterals and more fracture area lower the drawdown needed for a given rate, delaying the near-fracture pressure drop below the dew point. The second is how the wells are flowed: managed drawdown, holding back the early rate, keeps the near-well pressure higher for longer. Its benefit to ultimate condensate recovery in tight rock is debated, and the team will compare pilot wells flowed both ways. The third is target selection, landing wells in the parts of the Montney where the condensate-gas ratio and the economics are best. The last is the surface: high-pressure separation and stabilization to keep as many molecules as possible in the condensate product, and a deep-cut NGL plant for what remains in the gas.

Condensate markets

Condensate is valuable for different reasons in the two countries. In Canada, condensate is in high demand as diluent for bitumen from the oil sands, which must be thinned to flow in pipelines, and western Canadian producers have long imported condensate from the United States to meet that demand. Montney condensate, much of it produced in British Columbia and Alberta, is trucked or piped to the Edmonton area for that market, and the CER has noted that condensate from wet gas areas of the Montney is one of the formation's attractions. Either way, the liquids yield, measured in m³ per million m³, is one of the strongest drivers of the value of a well.