Page 473 - Enhanced Oil Recovery in Shale and Tight Reservoirs

Air injection 437

observed in the South Belridge field, California; the entire injection well-
bore temperature exceeded 538 C after 3 months of air injection (Gates

and Ramey, 1958). The oil viscosity was 2700 cP in the reservoir. In the pi-
lot test in the Holt Sand Unit, the reservoir temperature was increased from
200to 230 C (Fassihi et al., 2016). In 25 of the surveyed air injection pro-

jects, 5 projects had the evidence of spontaneous ignition (Chu, 1982).
However, Niu et al. (2011) proposed that the possibility of spontaneous
ignition was low; the thermal effect was minimum in light oil reservoirs, and
the air injection was considered immiscible flue gas injection. Greaves et al.
(1999) also believed that LTO, rather than in-situ combustion, prevailed in
air injection in light oil reservoirs; the important task was to study whether
the oxygen consumption rate was high enough so that the safety require-
ment could be ensured.

13.5.2 Laboratory observations

From their combustion tube tests, Montes et al. (2010) observed that the
GOR was maintained flat, indicating that the combustion front could miti-
gate gas viscous fingering, and the sweep efficiency was improved. Barzin
et al. (2010) conducted a series of ramped temperature oxidation tests for
light hydrocarbon oil and observed that when the temperature reached
about 180 C, the system suddenly increased about 60 C, indicating that

spontaneous ignition occurred through light oil oxidation.
However, Christopher (1995), and Yannimaras and Tiffin (1995) con-
ducted ARC tests using dozens of light oils and found that only 20% of
the oil samples had continuous exothermic behavior, and the rest of the
oil samples did not show the potential of spontaneous ignition.
Abu-Khamsin et al. (2001) conducted 22 air injection flooding tests us-
ing packed-bed reactors. They found that the maximum temperature in-
crease was about 10 C, indicating spontaneous ignition could not occur.

They interpreted that the LTO-generated heat was lower than the heat
loss. Jia et al. (2012a) conducted air flooding tests using their developed
apparatus and real cores to detect heat effect. Their system temperature
was maintained at a constant, and the reactor was wrapped using insulating
tapes to reduced heat loss. The core temperature was only increased from 80
to 89 C within 22.2 days of experiments. They also attributed such low

temperature increase to heat loss through metal thermocouples.
Clara et al. (2000) measured the temperature variation when air flowed
through their Handil cores under adiabatic conditions. Although the