SPE-185826-MS
Fluid Contact Monitoring Using API Tracking. A Novel Methodology for
Reservoir Surveillance
J. E. Paredes, A. Carbajal, R. Izquierdo, I. Loyo, R. Pérez, and L. M. Perera, Pemex E&P; C. J. Larez
Copyright 2017, SPE Europec featured at 79th EAGE Conference and Exhibition
This paper was prepared for presentation at the SPE Europec featured at 79th EAGE Conference and Exhibition held in Paris, France, 12–15 June 2017.
This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents
of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written
consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may
not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.
Abstract
There are different methods to estimate the current water-oil contact (WOC) or gas-oil (GOC) contact depth
in a reservoir. Reservoir engineers has commonly use PLT, RST logs, material balance and simulation
models with this purposed but they implied cost, man-hours and to have a great understating of the reservoir.
Previous published work1 proposed two diagnostic plots to estimate WOC or GOC depth at well level and to
estimate water or gas breakthrough using only measurement of °API, proving to be practical and accurate.
This work was based on the effect in the reservoir of a compositional gradient.
In this paper, a conceptual simulation model was built to prove the effect of the produced °API of
the presence of a compositional gradient in the reservoir. Results will show that distribution of °API
value in the reservoir will cause change, at surface level, in the measurement of this property during the
exploitation cycle, decreasing or increasing depending on weather the reservoir is undergoing a WOC or
GOC displacement. The results obtained with the conceptual simulation models present the same behavior
describe by other authors and could be translated to real models, and more important prove that changes
in the °APi values at surface level are a direct consequence of fluid displacement when the reservoir have
a compositional gradient.
Finally the methodology used in the conceptual simulation model for °API monitoring was applied to two
reservoirs, one with known aquifer influence and other with a secondary gas cap. Both reservoir had enough
historical °API measurement to establish a comparison between simulated and real data. The results show
that the numerical models are able to reproduce the historical °API behavior and reinforcing the practicality
of predict the breakthrough of fluids using only °API measurements on surface.
Introduction
There are different methods to estimate the water-oil (WOC) or gas-oil (GOC) contact in a reservoir. PLT
and RST allow phase monitoring within the reservoir in a timely manner, but they implies closing wells
while the logs are taken and differing production. Material balance has also been used to estimate current
fluid contact depth, however, it is necessary to have a porous volume vs. depth curve, rarely available
and when its available it has a high degree of uncertainty; Material balance analysis also assumes that the
displacement in the reservoir is uniform (piston like displacement).
2 SPE-185826-MS
Early work by Paredes et al1 discussed the usefulness of using surface measurements of °API and
proposed diagnostic plots, see Figure 1. These diagnostic plots can be used to estimate the depth of the fluid
contact and to predict the water and/or gas breakthrough into wells. The base of this two diagnostic plots
is the presence of a compositional gradient in the reservoir and its impact on the properties of the fluids
produced during reservoir exploitation.
Figure 1—Diagnostic plots proposed to estimate fluids contacts depth and fluids breakthrough.
The results confirm how the fluid displacement caused by the advance of the WOC or GOC generates a
decrease or increase of the °API and this changes can be used to estimate fluid contacts depth and/or predict
the time when water or gas breakthrough will occur in any producing well of the reservoir.
However, it is important to validate this phenomenon with a numerical model that sustains the issues
initially raised with this new methodology, eliminating any uncertainty associated with the real data
measurement. A conceptual model was built with the purpose of trying to reproduce and analyses the
behavior of °API and subsequently the methodology described in this paper was applied to real reservoir
models in order to confirm the described behavior. This conceptual simulation model will allow to
investigate the effect of compositional gradient and validate the hypothesis of the variation of the °API
because WOC or GOC movement in the reservoir.
Conceptual Reservoir Simulation Model
A single porosity, undersaturated and black oil model was built. The PVT analysis used had a proven
compositional gradient. The model has a 1,296 ft thickness to ensure the proper visualization of the effect
of compositional gradient in the reservoir.
ECLIPSE was choose for all the simulation to be run and the API tracking option2 was enable. The API
tracking in ECLIPSE allows to activate the mixing model of fluids with different °API in the simulation
model, although this study does not evaluated precisely this effect; by activating this option and adding a
compositional gradient in the model, it was possible to monitor the °API change in the simulation models
in different scenarios.
SPE-185826-MS 3
Characteristics of the conceptual model
For simplicity all petrophysical parameters (porosity, permeability, NTG, etc.) are constant. The model
includes a numeric aquifer in bottom to simulate the natural advancement of water in the reservoir. Table 1
shows a resume of all the main characteristics of the conceptual simulation model use in this paper.
Table 1—Conceptual simulation model characteristics.
Parameter Value
Porosity (%) 10
NTG (frac) 1
Horizontal permeability (mD) 100
Vertical permeability (mD) 10
Compressibility (1/psi) 4. 4885E-006
Top (m) 5002
Base (m) 5397
WOC (m) 5300
Temperature (F) 318.2
Initial pressure (psia) 8534
Bubble Pressure (psia) 3982
OOIIP (MMBls) 47.5
The model has a single producing well, located in the middle of the reservoir to allow the gradual
advancement of the water-oil contact and to better observe the phenomenon. Figure 2 shows a cross section
and top view of the simulation model.
Figure 2—Cross section and top view of the conceptual model.
Since it is a conceptual model, history matching wasn't required. Only the PVT data was matched to an
equation of State (EoS) and then the black oil tables were generated. Peng and Robinson of 3 parameters
was used as EoS and experimental data match was DL, CCE and separators experiments. Figure 3 shows
the final results of the matched EoS.
4 SPE-185826-MS
Figure 3—Matched EoS used for the conceptual simulation model.
Compositional gradient
In a high thickness reservoir, as the depth increases the amount of light components, the oil FVF and the RGA
values decreases while the amount of heavier components, the density and the bubble pressure increases3.
These changes in the properties of the fluid in the reservoir is due to compositional gradient. Sage and
Lacey4 were the first to experimentally observe the change in the composition of a mixture of hydrocarbons
with depth.
Compositional gradient is a direct consequence of the gravitational effects on the fluid in the reservoir,
the components of a mixture will migrate according to their mass, creating a variation of components
with respect to depth. Another factor of great importance to generate these compositional variations in the
reservoir are the changes of temperature and on a smaller scale, the phenomenon of molecular diffusion.
Compositional gradient has a greater effect in volatile oil or gas and condensate reservoirs, being almost
imperceptible in heavier oil reservoirs. Cases have been reported of compositional gradient in gas fields5.
In the case of the conceptual simulation model, the fluid gradient was generate after the EoS was matched
to the experimental data, this compositional gradient was added to the initialization profile in the form of
density vs depth and API vs depth because it was a black oil model and the change in composition can't be
added directly. This gradient obtained by using the matched EoS6 is shown in Figure 4.
SPE-185826-MS 5
Figure 4—Fluid gradient use in the conceptual simulating model.
Assuming that almost all compositional gradient models are considered isothermal or thermal, only
isothermal compositional gradient was evaluated in this paper, which is enough to consider the chemical
and gravitational effects within the reservoir of the compositional gradient. Figure 5 shows the conceptual
model initialization, showing the initial °API vs depth variation in the model. Although it wasn't monitored
in this study, the Figure 4 shows the effect of compositional gradient on the saturation pressure versus depth.
In this case the saturation pressure gradient is 0.067 bar/m, this value is within the expected range (0.025
bar/m for black oil and 1 bar/m for near critic fluids) proposed by Whitson5.
Figure 5—3D view of the °API initialization of the conceptual simulation model.
Results of the Model Concept
The results obtained in this study show that the behavior of the simulation model is consistent with what is
discussed in previous works about the effects of compositional gradient in the reservoir, specifically with
°API value. Normally heavier components (lower °API) are at the base of the reservoir and lighter (highest
°API) are located at the top of the structure. This distribution of °API value in the reservoir will cause
6 SPE-185826-MS
change, at surface level, in the measurement of this property during the exploitation cycle, decreasing or
increasing depending on weather the reservoir is undergoing a WOC or GOC displacement.
Two scenarios were analyzed using the conceptual simulation model:
▪ °API Monitoring in a Reservoir with WOC, Figure 6:
The objective of this scenario was to identify the effect on °API measurements of the WOC
displacement in the reservoir. The results shows that before the water breakthrough happens in the
producing well, the °API values in the simulation started increase slightly but as soon as the WOC
is near to the well, the °API values begin to decrease due to the displacement of the "heavier" oil
locate initially at the base of the reservoir and this trend will continue as the water cut increases.
Theoretically if the measurement of °API values has values closer to those initially measured at the
base of the reservoir (from the PVT analysis and compositional gradient) the well will present water
producing soon and the WOC is near to the well perforations.
▪ °API Monitoring in a Reservoir with GOC, Figure 7:
This scenario was use to analyze the effect on measurement of °API values when the GOC is
displaced in the reservoir. The results obtained with the simulation model confirms that the GOC
displacement in the reservoir move lighter components from the top of the reservoir to the producing
well and this makes a gradual increase in the °API values calculated by the simulation over time.
Taking into account this behavior, it will be possible to predict when the Pb will be reach in the
reservoir1, when the well will start producing gas from the gas cap and the location of the GOC in
the reservoir; using only °API measurement. For example, when °API measurements are closer to
the values measured initially at the top (from the PVT analysis and compositional gradient), the well
will start producing gas directly from the gas cap.
Figure 6—Conceptual simulation model results, showing °API - water cut profile and WOC - °API movement in the reservoir.
SPE-185826-MS 7
Figure 7—Conceptual simulation model results, showing °API - GOR profile and GOC-°API movement in the reservoir.
The result of the conceptual model, for both WOC and GOC displacement, confirms the hypotheses
raised in the works of Paredes et al1, which are summarized in figures 8 and 9:
1. In the WOC Case, the advance of the water-oil contact and the imbibition or drain effect in the reservoir
rock (depending on the wetting phase); the water front is displacing oil which was initially located
at the base of the reservoir, this progress will be reflected in a decrease of the °API value measured
in production wells.
2. When the oil displacement is caused by the influence of the gas cap (GOC case), the effect is reverse;
the °API values measured at surface level will begin to increase as the lighter oil is displaced towards
the producing well.
8 SPE-185826-MS
Figure 8—Outline showing the expected °API and water cut behavior in a reservoir with compositional gradient. (SPE 185588).
Figure 9—Real well data showing the °API behavior (SPE 185588).
Practical Application
The conceptual model proved that it is possible using a simulation model to monitor °API variations with
respect to the water cut and gas-oil ratio. From the point of view of numerical simulation, it is not necessary
to monitor the °API to predict the fluids contacts movement within the reservoir. This conceptual models
allowed to confirm the assumptions associated with the presence of a compositional gradient in the reservoir
and its effect on °API variation during the reservoir exploitation life cycle and, thus to give confidence that
the water or gas breakthrough time can be predicted using only measurements of °API at surface level1.
Similar to the conceptual simulation model, the °API monitoring methodology for simulation model was
used in two fields that already had simulation models previously history matched and with enough real
measurements of °API to make a validation between the results of the simulation model and the actual
measurements possible. The characteristics of both simulation models are resuming in the Table 2.
Table 2—Characteristic of Reservoir A and B simulation models.
Reservoir A (WOC Case) Reservoir B (GOC Case)
Model type BO / DP Model type BO / DP
Number of Cell 190 M Cell no. 130 M
WOC (ft) 18,340 / 17,355 / 17,060 WOC (ft) 18,274
Initial Pressure (psig) 7,570 Initial Pressure (psig) 14,100
SPE-185826-MS 9
Reservoir A (WOC Case) Reservoir B (GOC Case)
Model type BO / DP Model type BO / DP
Bubble pressure (psig) 4.926 Bubble pressure (psig) 3,342
OOIP (MMb) 147 OOIP (MMb) 286
Wells 16 Wells 7
Results Reservoir A
The selected field is a light oil naturally fractured reservoir with different WOC and with a high bottom
aquifer influence. Nowadays, all wells in this reservoir had presented water breakthrough, making it the
ideal candidate for testing the methodology presented in this work for the advancement of WOC. Figure
10 presents a 3D section showing the saturation of fluid at the end of history and the final history match
of the field.
Figure 10— 3D section and history matching (Reservoir A).
The results obtained indicate that effectively °API values decreases in all wells as the water cut increases.
Figures 11 and 12 present the results for one particular well within the model, that describe best this behavior.
Figure 11—Comparison of surface °API data (observed vs simulated) and water cut for "Well A" (Reservoir A).
10 SPE-185826-MS
Figure 12—3D section showing °API behavior in 3 time steps (Reservoir A).
Clearly the simulation model was able to reproduce °API behavior and that this behavior is exactly what
was described in this paper. These findings strengthen the theories presented and give more reliability to
the use of °API measurements to predict the location of fluid contacts and breakthrough times.
Results Reservoir B
This reservoir has a volumetric behavior (without aquifer influence) and the only well that presented water
production is located on the base of the structure. Figure 13 shows the history matching of the field model
and a 3D section showing the current fluid saturation of the field.
Figure 13—3D section and history matching (Reservoir B).
This field was selected as the Pb was achieved and was able to form a secondary gas cap, making it
a candidate for excellence to test GOC advance using surface °API information. Applying API Tracking
process was possible to monitor the surface °API behavior in time. Figure 14 presents the historical variation
of °API and a section of the reservoir for 3 time steps where it can be shown how is displaced the lighter
oil to the producer well.
SPE-185826-MS 11
Figure 14—Behavior of the surface API for 3 time steps (Reservoir B).
Comparing simulated results with field surface °API data for "Well A" can appreciate how is manages
to reproduce the behavior of °API and GOR, Figure 15. This behavior confirms the initial approach of the
variation of the °API in the reservoir.
Figure 15—Comparison of surface °API data (observed vs simulated) for "Well A" on the Reservoir B.
Conclusions
This work confirms that, using surface °API data is possible to monitor the WOC and GOC advances and
furthermore predict breakthrough time without relying on a numerical model.
Using a conceptual numerical simulation model it was possible to demonstrate the effect of water or
gas advance on the behavior of the API within the reservoir; this effect is associated to the presence of a
compositional gradient in the reservoir.
Applying API tracking option it was possible to reproduce in real field models, the decrease of the
°API due to WOC advance and the increase by the influence of the gas cap, strengthening the proposed
methodology.
It is shown using numerical models the effects of the presence of a compositional gradient in the °API
of the fluid produced, validating the results presented in previous paper1 and reinforcing the practicality of
predict the breakthrough of fluids using only °API measurements on surface.
Nomenclature
API Unit of Specific gravity of oil
12 SPE-185826-MS
WOC Water-oil contact
GOC Gas-oil contact
EoS Equation of State
FVF Formation Volume Factor
NTG Net to Gross
Pb Bubble point pressure
PLT Production Log Tool
PVT Pressure, Volume and Temperature study
Pr Reservoir pressure
GOR Gas-oil ratio
RST Reservoir Saturation Tool
Tr Reservoir temperature
References
1. J.E. Paredes et al., "Methodology for Fluid Contact Monitoring through Surface API Data", SPE
185588, Buenos Aires 2017.
2. SCHLUMBEGER ECLIPSE Reference MANUAL.
3. Mokhtari Rasoul, and Siavahs Ashoori. "Importance of compositional grading in reservoir
development studies; a case study" ISSN 1013-5316, Sci.Int, 2013.
4. Sage, B.H., Hicks, B.L., and Lacey, W.N.: "Partial Volumetric Behavior of the Lighter
Hydrocarbons in the Liquid Phase" Drill. & Prod. Prac., 1938.
5. Whitson C H and Belery P., "Compositional gradients in petroleum reservoirs", SPE 28000
presented at University of Tulsa Centennial Petroleum Engineering Symposium, Tulsa 1994
6. Pedersen K.S., Lindeloff N., "Simulation of Compositional Gradients in Hydrocarbon Reservoirs
Under the Influence of a Temperature Gradient", SPE 84364, Denver 2003.
7. Whitson C.H., Michael R. Brulé: "Phase Behavior", SPE Monograph Series Volume 20,
Richardson-Texas, 2000.
8. McCain W.D.: "The Properties of Petroleum Fluids", PennWell Publishing Company, Tulsa-
Oklahoma, 1990
