Dear FeFlow-community,
I have several issues when running FeFlow-ATES-simulations.
Problems in first study case: I have built a model containing a reservoir
consisting of two lithologies with varying petrophysical properties. 80% of this
reservoir (ambient T: 13°C) is covered by the well screens. This model contains
5 wells (ideal cell size considered), where 3 wells represent the cold side
(injection with ~40°C, varying flow and production rates, and a distance of
~200m between these wells) and 2 wells the warm side (injection with ~90°C, also
varying flow and production rates, and a distance of ~250m; distance to cold
wells: ~350 m). In the first case ("model 1"), I have a hydraulic gradient from
North to South, in the second case ("model 2"), I have a constant water level
(hydraulic head boundary at all sides of the model). In all cases, I use the
SAMG Equation-System Solver SAMG. Screenshots of settings in figures 1-3
attached.
1. Problem: In Model 1, I observe the effect that during an extraction phase,
one of the warm wells and one of the cold wells (always the same pair during a
single run) exhibit a completely different behavior and temperature range (see
figure 4). However, this effect seems somewhat arbitrary because it affects
different wells from run to run, hence I don't think that the distance between
the wells are the reason for this effect. Additionally, when I include more
wells in the model, it sometimes happens that only one cold well is affected by
this effect, while none of the warm wells are. At this point, I am very confused
because I can’t identify any pattern and am certain that I have parameterized
all the boreholes identically. Moreover, the only thing I’ve varied in the model
is the number of wells (and hence the flow rates at each well), nothing else…
2. Problem: In Model 2, I don’t observe the effect described earlier. However, a
different issue occurs here. The temperature at the wells remains absolutely
constant during the extraction phases, both on the warm (Fig. 5) and cold sides
(Fig. 6). Essentially, I get a step-like diagram. Additionally, the
temperatures, particularly on the cold side, are extremely low, as if the
injection of warm water has no effect on the subsurface. Moreover, the
extraction temperature on the warm side also fluctuates unusually strongly, from
one cycle to the next. As with Model 1, I have parameterized this model in the
exact same way, except for using a different hydraulic head boundary.
In this model, I set up one cold and one warm well spaced approximately ~1500
meters apart. The reservoir has an ambient temperature of around 55°C, with warm
water (90°C) and cold water (40°C) being injected at varying flow rates.
The main issue arises when comparing scenarios with a hydraulic gradient versus
a constant water level. The simulation with a hydraulic gradient shows
significantly higher heat recovery efficiencies. This difference is primarily
due to variations in extraction temperatures:
- With a hydraulic gradient: The extraction temperature on the warm side starts
higher and decreases linearly. On the cold side, the extraction temperature
starts below the injection temperatures of 40°C and also increases linearly (see
figure 7).
- Without a hydraulic gradient: The warm side's extraction temperature decreases
more gradually, reaching lower values over time. On the cold side, the
extraction temperature initially exceeds the injected temperature but slows down
as the extraction cycle progresses (see figure 8).
These differences significantly impact the efficiency of heat recovery under
varying conditions.
Has anyone observed similar effects and found a explanation/solution for them? I
would be highly thankful for any help or tipp!