def test_swcrsorw():
"""Test estimate_sorw/swcr for some manually set up cases"""
# swcr, sorw, h:
testtuples = [
(0, 0.3, 0.1),
(0.01, 0.2, 0.1),
(0.3, 0.1, 0.1),
(0.0, 0.0, 0.1),
(0.00001, 0.9, 0.1),
(0.0001, 0.99, 0.1),
]
for testtuple in testtuples:
real_swcr = testtuple[0]
real_sorw = testtuple[1]
h = testtuple[2]
wateroil = WaterOil(swcr=real_swcr, sorw=real_sorw, h=h)
wateroil.add_corey_oil(now=2, kroend=0.8)
wateroil.add_corey_water(nw=2, krwend=0.9)
est_sorw = wateroil.estimate_sorw()
mis = abs(est_sorw - real_sorw)
print("Testing sorw={}, swcr={} on h={}".format(real_sorw, real_swcr, h))
if mis > 0.01:
print("Missed, estimated was {}".format(est_sorw))
assert mis < h + epsilon # Can't guarantee better than h.
est_swcr = wateroil.estimate_swcr()
mis_swcr = abs(est_swcr - real_swcr)
if mis_swcr > 0.0:
print("Missed swcr, estimate was {}".format(est_swcr))
assert mis_swcr < h + epsilon
def test_normalize_nonlinpart_wo():
"""Manual tests for utils.normalize_nonlinpart_wo"""
wateroil = WaterOil(swl=0.1, swcr=0.12, sorw=0.05, h=0.05)
wateroil.add_corey_water(nw=2.1, krwend=0.9)
wateroil.add_corey_oil(now=3, kroend=0.8)
krwn, kron = utils.normalize_nonlinpart_wo(wateroil)
assert np.isclose(krwn(0), 0)
assert np.isclose(krwn(1), 0.9)
# kron is normalized on son
assert np.isclose(kron(0), 0)
assert np.isclose(kron(1), 0.8)
# Test with tricky endpoints
h = 0.01
wateroil = WaterOil(swl=h, swcr=h, sorw=h, h=h)
wateroil.add_corey_water(nw=2.1, krwend=0.9)
wateroil.add_corey_oil(now=3, kroend=0.8)
krwn, kron = utils.normalize_nonlinpart_wo(wateroil)
assert np.isclose(krwn(0), 0.0)
assert np.isclose(krwn(1), 0.9)
assert np.isclose(kron(0), 0)
assert np.isclose(kron(1), 0.8)
# Test again with zero endpoints:
wateroil = WaterOil(swl=0, swcr=0, sorw=0, h=0.01)
wateroil.add_corey_water(nw=2.1, krwend=0.9)
wateroil.add_corey_oil(now=3, kroend=0.8)
krwn, kron = utils.normalize_nonlinpart_wo(wateroil)
assert np.isclose(krwn(0), 0.0)
assert np.isclose(krwn(1), 0.9)
assert np.isclose(kron(0), 0)
assert np.isclose(kron(1), 0.8)
# Test when endpoints are messed up:
wateroil = WaterOil(swl=0.1, swcr=0.2, sorw=0.1, h=0.1)
wateroil.add_corey_water(nw=2.1, krwend=0.6)
wateroil.add_corey_oil(now=3, kroend=0.8)
wateroil.swl = 0
wateroil.swcr = 0
wateroil.sorw = 0
krwn, kron = utils.normalize_nonlinpart_wo(wateroil)
# These go well still, since we are at zero
assert np.isclose(krwn(0), 0.0)
assert np.isclose(kron(0), 0)
# These do not match when endpoints are wrong
assert not np.isclose(krwn(1), 0.6)
assert not np.isclose(kron(1), 0.8)
# So fix endpoints!
wateroil.swl = wateroil.table["sw"].min()
wateroil.swcr = wateroil.estimate_swcr()
wateroil.sorw = wateroil.estimate_sorw()
# Try again
krwn, kron = utils.normalize_nonlinpart_wo(wateroil)
assert np.isclose(krwn(0), 0.0)
assert np.isclose(kron(0), 0)
assert np.isclose(krwn(1), 0.6)
assert np.isclose(kron(1), 0.8)
def test_swcr():
wo = WaterOil(swcr=0.3, swl=0.1, sorw=0.1, h=0.05)
wo.add_corey_oil(now=2, kroend=0.8)
wo.add_corey_water(nw=2, krwend=0.6)
est_sorw = wo.estimate_sorw()
est_swcr = wo.estimate_swcr()
assert np.isclose(est_sorw, 0.1)
assert np.isclose(est_swcr, 0.3)
def test_swcr():
"""Test that we can locate swcr with a certain accuracy"""
wateroil = WaterOil(swcr=0.3, swl=0.1, sorw=0.1, h=0.05)
wateroil.add_corey_oil(now=2, kroend=0.8)
wateroil.add_corey_water(nw=2, krwend=0.6)
est_sorw = wateroil.estimate_sorw()
est_swcr = wateroil.estimate_swcr()
assert np.isclose(est_sorw, 0.1)
assert np.isclose(est_swcr, 0.3)
def test_interpolations_wo_fromtable():
"""Analog test as test_interpolations_go_fromtable().
Pyscal 0.6.1 and earlier fails this test sorw.
"""
base = pd.DataFrame(
columns=["Sw", "krw", "krow"],
data=[
[0.0, 0.0, 1.0],
[0.1, 0.0, 1.0],
[0.2, 0.0, 1.0], # swcr
[0.3, 0.1, 0.9],
[0.8, 0.8, 0.0], # sorw
[0.9, 0.9, 0.0],
[1.0, 1.0, 0.0],
],
)
opt = pd.DataFrame(
columns=["Sw", "krw", "krow"],
data=[
[0.0, 0.0, 1.0],
[0.1, 0.0, 1.0],
[0.3, 0.0, 1.0],
[0.4, 0.1, 0.2], # swcr
[0.9, 0.9, 0.0], # sorw
[0.95, 0.95, 0.0],
[1.0, 1.0, 0.0],
],
)
wo_base = WaterOil(h=0.01)
wo_base.add_fromtable(base)
assert np.isclose(wo_base.estimate_swcr(), 0.2)
assert np.isclose(wo_base.estimate_sorw(), 0.2)
wo_opt = WaterOil(h=0.01)
wo_opt.add_fromtable(opt)
assert np.isclose(wo_opt.estimate_swcr(), 0.3)
assert np.isclose(wo_opt.estimate_sorw(), 0.1)
wo_ip = interpolate_wo(wo_base, wo_opt, 0.5, h=0.01)
assert np.isclose(wo_ip.estimate_swcr(), 0.25)
assert np.isclose(wo_ip.estimate_sorw(), 0.15)
def test_linear_corey():
"""Test how the estimate_sorw works when Corey is linear"""
# Piecewise linear, this we can detect:
wateroil = WaterOil(swl=0, h=0.01, sorw=0.3)
wateroil.add_corey_water(nw=1, krwend=0.8)
est_sorw = wateroil.estimate_sorw()
assert np.isclose(est_sorw, 0.3)
# Piecewise linear, this we can detect:
wateroil = WaterOil(swl=0, h=0.01, sorw=0.5)
wateroil.add_corey_water(nw=1, krwend=1)
est_sorw = wateroil.estimate_sorw()
assert np.isclose(est_sorw, 0.5)
# This one is impossible to deduct sorw from, because
# the entire krw curve is linear.
wateroil = WaterOil(swl=0, h=0.01, sorw=0.5)
wateroil.add_corey_water(nw=1, krwend=0.5)
wateroil.add_corey_oil(now=2)
est_sorw = wateroil.estimate_sorw()
# The code will return sw = swl + h by design:
# Unless information is in the krow table, anything better is impossible
assert np.isclose(est_sorw, 1 - 0.01)
def test_sorw_hypo(h, sorw, nw):
"""Test estimate_sorw extensively"""
wateroil = WaterOil(sorw=sorw, h=h)
wateroil.add_corey_oil(now=2)
wateroil.add_corey_water(nw=nw)
est_sorw = wateroil.estimate_sorw()
est_error = abs(sorw - est_sorw)
assert not np.isnan(est_sorw)
error_requirement = h + 10000 * epsilon
if abs(1 - nw) < 0.1:
# When the Corey curve is almost linear, we don't
# expect sorw estimation to be robust
error_requirement = 3 * h
if est_error > error_requirement:
print("h = {}, sorw = {}".format(h, sorw))
print("estimated sorw = {}".format(est_sorw))
print(wateroil.table)
# We don't bother to tune the code better than this:
assert est_error <= error_requirement
