SPE 59471

Historical Problems with Old Field Rejuvenation

F.T. Blaskovich,SPE, Blaskovich Services, Inc.

Copyright 2000, Society of Petroleum Engineers Inc. twentieth century was rightly called the Oil Century. The real
This paper was prepared for presentation at the 2000 SPE Asia Pacific Conference on oil production history of the planet actually dates back to 3000
Integrated Modelling for Asset Management held in Yokohama, Japan, 25–26 April 2000.
B.C., but nobody then cared much about accurately measuring
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, as
the volumes as it seeped out of the ground in various parts of
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to the Middle East, Europe and elsewhere 1 .
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at More than 100 billion metric tons (800 billion stock tank
SPE meetings are subject to publication review by Editorial Committees of the Society of barrels) have been produced in the last 150 years 2 . According
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is to recent estimates 3 (see Table 1), this represents less than
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 half of the estimated total world recoverable oil reserves.
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. There are also estimates that another 40 to 140 billion tons are
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
yet to be discovered 4 . In any case, with a steadily increasing
demand projected, it is now expected that worldwide oil
Abstract production will start to decline in the next 5-20 years,
In many parts of the world, the redevelopment of older fields depending on how optimistic the forecast is.
presents unique problems that will require innovative Oil field technology has changed dramatically in recent
solutions. With the rising costs of finding and developing new time. Since the 1970’s, modern computer technology has
oil fields, oil companies are turning more attention to helped to improve our understanding of reservoir geology and
revitalizing existing fields. Because of historical practices, fluid flow. Field measurement techniques have also greatly
these fields often have significant potential for enhanced improved. Much of this coincided with the needs of the
recovery. These older fields have many different problems to harsher operating environments of the North Sea and Alaskan
deal with than the more modern fields that traditionally drive North Slope, which started development at that time.
petroleum technology. Today, technology enables us to develop, operate and
Advances in computer hardware and software technology understand difficult fields and produce them more
have made it possible to analyze reservoirs at a level of detail economically than would have been possible in the past.
unimagined only a few years ago. With all of this power on Modern measuring and analysis tools reduce the uncertainties
our desktops, it is easy to lose sight of a major problem in in field development and allow operators to control costs
dealing with older fields – the historical data. effectively.
In many cases, there may be a large amount of historical
data, but much of it is either inaccurate or critical types of data Old Fields – Of the remaining already discovered oil (148
have not been gathered. Several field examples are discussed billion tons), 80% of it is contained in the Middle East, the
in this paper. Each field has redevelopment potential using Former Soviet Union (FSU), and South America where many
modern techniques, but each project has great uncertainty due fields can be classified as “old” fields. For example, the
to data constraints. The conclusions and recommendations of supergiant Ghawar field in Saudi Arabia and the Burgan field
the paper include suggestions for ways to solve these in Kuwait were found in the late 1930’s. Major fields in Iran
problems. Since we cannot go back in time to gather critical and Iraq were found even earlier. The Siberian oilfields in
missing data (e.g., early production and injection data, initial Russia were found in the 1950’s. The Maricaibo basin fields
fluid properties, produced compositions, etc.), special in Venezuela started in the 1920’s.
techniques will be needed to determine the value of these Most of the remaining discovered oil in the world is either
projects. in fields that were initially developed before modern
technology was available (pre-1970) or in fields that have
Introduction been developed more recently without the use of modern
We will soon mark the 150th anniversary of the discovery of philosophies and technologies. Therefore, “old” is much more
commercial quantities of oil in Pennsylvania and the start of a function of development and operating practices than time.
the modern petroleum industry. Since that time, oil has truly Figure 1 shows an estimate of the remaining discovered oil as
been the driver of industrialization around the world. The
2 F.T. BLASKOVICH SPE 59471

a function of discovery time. More than half of this oil (or 80 we can do is to understand that errors will exist and try to
billion tons) is in these “old” fields. This creates a unique and quantify them as best we can in the reservoir analysis process.
interesting problem in the evaluation of these fields for
potential rejuvenation. Old fields generally have a lot of The Problem Defined
historical data (production and injection volumes, well Historical data errors are more common than we realize or
pressures, etc.), but this data can often be very inaccurate. want to admit. They can arise from human error or neglect,
Traditionally, errors in the pressure, production and mechanical malfunctions, infrequent measurements when field
injection data have been ignored in the petroleum literature. conditions are changing, imprecisely calibrated tools, or a lack
In recent years, there has been increasing emphasis on risk and of enough detail in the measurement procedure itself. If large
uncertainty analysis in field performance studies, but it has enough and widespread, they can make a difference in our
generally been applied to the geologic description 5 , the understanding of the potential of a field.
material balance 6-10, and/or to production forecasts 11,12 . In
each case, risk analysis is justified because we have very Pressure Data Errors. Bottom-hole pressure measurements
limited information about the properties we are trying to are a key ingredient in understanding the dynamic behavior of
describe. In old fields, we have the additional complication of an oil reservoir. We rely on them to be accurate in our models
having a historical dataset with an unknown degree of so that fluid properties will be correctly used and also to
accuracy. All reservoir engineering analysis techniques make provide an understanding of the connected pore volume, fluid
the assumption that the pressure-production data is accurate. transmissibility and aquifer strength. The pressure data is vital
Therefore, a different method is needed if there is reason to to correctly interpret the material balance and wellbore flow.
believe that this is not the case. The most common errors in pressure data arise from (1)
The problem of inaccurate pressure, production or tool malfunctions and improper calibration, (2) downhole
injection measurements is seldom mentioned in the literature. communication through tubing and casing leaks caused by old
Perhaps this is because most field re-development has been or sub-standard equipment, (3) a lack of knowledge of the
done where data errors were not a problem in fields with average wellbore fluid gradient and the real depth of the
reliable historical databases. Maybe the errors were small and pressure gauge, and (4) the general lack of measurements at
did not affect the solution. Maybe no techniques were readily any useful frequency to allow us to understand the relationship
available to deal with the problem. between pressure and fluid flow. Most of these errors do not
In the last few years, there has been a tremendous push by exist in today’s modern high tech fields where continuous
major oil companies to participate in the redevelopment of monitoring and measurement is more often the rule.
fields previously operated by others. Good examples of this
are the U.S. and European oil companies rushing into Kuwait, Production Data Errors – In general, total field oil
Russia, Venezuela and China during the 1990s. Evidence of production is reliable if the oil is metered and sold. Data
the historical data problems was well documented in recent reliability will vary depending on when and where it was
papers 13,14 about the redevelopment of the Burgan field in obtained. Oil data has more meaning to the operating
Kuwait, which is the second largest oilfield in the world. company and its staff and is usually handled more carefully.
Errors in the pressure, production and injection data can Errors in the total oil volume can still occur through improper
create significant errors in reservoir simulation models unless meter calibration, measurement errors, or facility and flowline
they are acknowledged and accounted for during the history leaks which reduces the amount of produced fluids that are
matching process. Risk analysis techniques can be used to metered. In contrast, total water and gas production may not
bracket the range of these data errors and provide valuable be accurately measured or even reported. It is common in old
insight for the reservoir engineer and re-development team. fields to report a “number” because it is required, but not to
An intelligent assessment can reduce the possibility that a non- worry too much about its accuracy.
physical reservoir description will be created by trying to While total field oil production is directly tied to
match unrealistic datapoints. This, in turn, can reduce the risk profitability, quotas, salaries and bonuses, oil production at the
that prediction results will be unreliable. individual wells is much less important. In most older fields,
By itself, an understanding of historical data errors is not this data is inferred by comparing infrequent well test rates to
enough to guarantee reliable simulation models. Reservoir the total metered oil production. Individual well rates are
description uncertainties and errors will always be an determined by allocation. This can easily create errors in the
important driver of the model results. An understanding of production database through test errors and especially if
both types of data problems is needed for a successful effort. downtime is not correctly accounted for or wells share a
The development of new technologies will not reduce or common flowline.
eliminate the historical data problems. Once this data is Gas and water production at the well level can easily be
gathered, we cannot measure it again with better tools or with corrupted by a combination of faulty field level measurements,
more care. The damage has been done and cannot be undone, allocation problems, and poor test separator measurements.
unless we invent a time machine and travel back to the period For accurate reservoir models, it is necessary to know the
in question. Since that is unlikely in the near future, the best oil, gas and water production and injection at an even more
detailed level. Ideally, volumes should be reported at the
SPE 59471 HISTORICAL PROBLEMS WITH OLD FIELD REJUVENATION 3

reservoir and perforation levels. This requires even more data However, it is possible to bracket the range of likely answers
that can be flawed by errors. In a multi-reservoir field, it is and, hopefully, make an intelligent decision about the
likely that, over time, at least some wells will experience acquisition or operation of the field.
wellbore communication. This may be very difficult to detect
if the reservoirs have similar properties. Within a single Examples
reservoir, multiple perforation intervals are likely. Normally, The following are examples showing some of the problems
we rely on production logs to decipher how much fluid came described above occurring in real fields. These problems are
from each interval. It is likely in old fields that these logs are not confined to any particular region of the world or time
not available for most or all of the historical period. period. The chances of these types of problems occurring
greatly increase as the age of the field increases. However,
Injection Data Errors – As with gas and water production even fields developed in the last few years share many of these
data, this data can be wrong for many reasons. In some fields, problems, as is shown below.
the injection volumes that are reported may have little to do
with the actual injection rates. Allocation based on well tests Example 1 – FSU – This field was discovered in the FSU in
is the most popular way to fill the database. This means that 1985 and put on production in 1987. Only one reservoir was
errors in metering at the field or gathering center level present. The oil was initially highly undersaturated and of
combined with inaccurate well test measurements can moderate to high API gravity. Before development began, it
significantly affect the reported results. Missing data in old was decided that a fieldwide water injection program should
fields can also create uncertainty. Leaks in flowlines and in be started as soon as possible, based on performance data from
the tubing and casing can also cause a mismatch between surrounding fields. This proved to be a good decision because
metered and injected volumes. There is also the possibility the reservoir pressure started to decline almost immediately
that fluids can be injected into the wrong reservoirs. after production started.
Water injection began in 1988. The injection program was
Ways Forward - The key issue is to recognize that any or all followed as part of the original development plan. Cumulative
of these errors can occur and to build contingencies into the injection volumes that were reported in the historical data
historical database for modelling. First of all, the available exceeded reservoir voidage by twenty percent by the time it
data should be scrutinized to see if any obvious anomalies was stopped in 1995.
exist. For example, is there data reported for all wells at all Using the reported data, a simulation model was built and
reporting times between when the well was drilled and matched. This “deterministic” history match for Field 1 is
completed and final shut-in? Often the final database of shown on Figures 2 through 4. At first glance, the match
values is all that is available. It may be very difficult to check appears to be excellent. On a fieldwide basis, the historical
reported volumes against metered volumes. data and the model agree closely. However, a closer
It is always worthwhile to talk to the people operating the inspection of the well matches is shown on Figures 5 and 6 for
field and find out how they measure fluids and pressures. In producers and injectors, respectively. These plots show the
addition, it is very useful to learn as much as possible about quality of the well matches. Each dot represents a well. The
the social, economic, and historical background of the region x-coordinate of each dot represents the well’s percent
to get a better understanding of how field operations may have deviation from the reported production (or injection) volumes
been impacted. and its y-coordinate is a measure of the relative importance of
For questions about well rate allocation, it is useful to have the well. Relative importance is determined as the ratio of the
a flowline diagram and some knowledge of the historical well’s production (or injection) to the total field production (or
surface facility configuration. With this data, it may be injection). As shown, most of the well matches are not
possible to recreate the surface operations with a pipeline satisfactory.
model and check the appropriateness of the split between Using the “deterministic” water injection data in the
wells. Much of the time the data needed is not readily simulation model, it was necessary to increase permeability in
available. Even so, some simple calculations may be useful to large areas around the water injectors by more than an order of
bracket the range of possible well rates. magnitude to match the injection volumes. This was much
In many cases, more creative methods need to be adopted. higher permeability than was originally assigned based on
Using a risk analysis program coupled to a spreadsheet, it is core and build-up test data. If the water injection values are
possible to assign a possible range of flowrates to each well in correct, it is likely that oil was displaced out of the structure.
the field. These ranges then become part of the history This represents one possible explanation of the field history,
matching parameters. Instead of only adjusting the reservoir but one that doesn’t fit all of the available data.
description during the match, it is possible to adjust the Another possibility is that the pressure, production and/or
production and injection rates as well. This becomes injection volumes were in error. A closer examination of the
particularly valuable when artificial adjustments to the database received from the operator revealed several
reservoir description are needed to explain production or interesting facts. The pressure database did not extend over
injection. (See Example 1 below.) It must be remembered the life of the field. In months where pressures were recorded,
that, in these cases, it is impossible to get an accurate answer. it was often the case that only a few wells were tested. In
4 F.T. BLASKOVICH SPE 59471

addition, the pressure gauge depth was not available for much or not. If it did inject, the range of possible values is quite
of the data. Production and injection data was missing for large since the well had historically been a good injector. This
many wells during the water injection period in the mid- explains the wide range of possible values shown between
1990s. Evidence from the field indicated that flowline leaks 1992 and 1994.
had occurred. All metering was done at the field level and
well rates were allocated on the basis of infrequent tests. The Example 2 – Western Siberia – This field was discovered in
field was in a harsh arctic environment and operational the mid-1950s and brought on production at that time. The
problems had occurred. The field was an offset to a much field consists of a series of sand-shale sequences with as many
larger field and had been neglected to some extent. as 10 producing horizons. Development began in the deeper
Using the information available, it was possible to create a reservoirs first and gradually spread to the upper sands. Water
new pressure, production and injection history. Error bars injection began shortly after production started. As with Field
were assumed and tested during the history matching to ensure 1 described above, more water was injected than fluids
consistency and reasonableness. The revised history data was produced and several of the majors sands were overpressured.
generated using a Monte Carlo program and a spreadsheet. Based on the available production and injection data, a
Factors used included: deterministic history match was performed. The total field
production and injection matches are shown on Figures 13
• Field metering error (+/- 5%) through 15. While this is a good overall match, the individual
• Well test measurement error (+/- 5%) well matches are poor, as shown on Figures 16 and 17.
• Leak occurrence rate (25%) Table 2 shows a summary of the main characteristics of
• Leak loss rate (5-25%) the production and injection database for Field 2. As shown,
there are many discrepancies in the data. There is a large
These values were chosen to illustrate the method and fit amount of uncertainty concerning which reservoirs wells were
the observed data. In reality, we do not know the actual completed in, whether some wells produced from more than
values and never will. These events happened in the past and one reservoir at the same time, and some wells that had
poor or missing data make them uncertain. They must be production or injection, but no perforations, etc.
inferred from local knowledge and should be adjusted if In addition to these problems, the field would also be a
results become unreasonable. likely candidate for the same problems experienced in Field 1,
To demonstrate the impact of these potential errors, the such as measurement errors, flowline and downhole leaks, and
pressure, production and injection data with potential error possibly interference caused by multiple wells sharing the
bars were plotted for Field 1. same flowline. The historical data is further complicated by
Figure 7 shows the average field pressure based on the fact that the field has many reservoirs and many wells
observed data. The length of each vertical black line is a penetrate most of them. Tubing and casing leaks and poor
measure of the uncertainty in the field average pressure. cement jobs are very common in this part of the world and this
Figure 8 shows a typical well’s pressure history with makes it likely that there has been a significant amount of
corresponding error bars. unintentional downhole commingling during the life of the
Figure 9 shows the error bars for the total oil production field.
rate. The green shaded areas around the mean (shown by a As with Field 1, the next step in the history matching
black line) represent one standard deviation on either side of process would be to re-analyze the historical database and
the mean. The red shaded areas extend the potential values develop a more “probabilistic” version before attempting any
out to the 5th and 95th percentile. Therefore, all but 10 percent further model runs.
of the possible values for oil production lie in the shaded
areas. All plots shown in this paper use these same criteria. Example 3 – South America – This field was discovered in
Figure 10 shows a typical production well. The plot at the 1927 and was quickly developed during those early days when
bottom of the figure is a measure of data availability. Where oil technology was in its infancy. Drilling was done quickly
blank spots occur, there is missing data. Correspondingly, the and logging was haphazard. The goal was to get as many
range of possible values in the production plot increases wells producing oil as quickly as possible. There was little
during those periods. There is more flexibility in those regard for measuring pressures or water and gas production.
periods due to increased uncertainty. For this particular well, Water injection for secondary recovery or pressure
there were no blank spots in the production database. maintenance was just starting in the U.S. in 1925 15 and did
Figure 11 shows the field wide water injection history. not become a common practice until the 1950s.
Note that the range of uncertainty increases during the middle Field 3 could have benefited from early water injection.
years. This is the period when much of the production and The field is composed of many highly stratified reservoir
injection data was missing from the database. layers and is highly faulted. The result is that many portions
Figure 12 shows a typical water injector in Field 1. In this of the field are in isolated fault blocks and have limited
case, there are significant gaps in the injection database. communication with the regional aquifer. Because pressure
During these periods, it is uncertain if the well injected water maintenance was not used in the early years, the reservoir
pressure and well productivity declined rapidly. A secondary
SPE 59471 HISTORICAL PROBLEMS WITH OLD FIELD REJUVENATION 5

gas cap was formed in the crestal structure and gas-oil ratios quality of all available data before a reliable analysis can
decreased initially as free gas migrated updip and then be made.
increased as this gas was produced. 2. A sense of the social, economic and historical setting in
Figure 18 shows the production history of Field 3. There which the field was developed is important for
are many problems with this data. Note that the first twelve understanding the type and magnitude of potential data
(12) years of the production history are missing. This could errors that can occur.
be a major problem because there is good reason to believe 3. Uncertainty techniques, such as Monte Carlo simulation,
that this was a period of very high rate production. All other need to be used to re-condition and understand the
fields in this region that were on production at or near this potential variations in historical input data for simulation
time were producing at their highest rates. models.
Note that the gas-oil ratio during the period before 1955 is
constant at 242 m3 /ton. It is possible that this did occur, but it References
is unlikely with the pressure dropping during this lengthy 1. Yergin, D.: The Prize: The Epic Quest for Oil, Money,
period. Another explanation is that the gas production report and Power, Simon and Schuster, New York, N.Y. (1991).
is not accurate before 1955. Gas was not an important 2. Petroleum: An Energy Profile 1999, Energy Information
commodity at the time and was worth very little compared to Administration, Washington, D.C. (1999).
the oil. Most of it was probably flared. 3. International Petroleum Encyclopedia, OGJ (1999).
Field 3 is a good example of the nature of very old fields. 4. Laherrere, J: “World Oil Reserves: Which Numbers to
It seems that the older the field, the more likely it is that Believe?”, OPEC Bulletin (Feb. 1995), 9.
historical data errors will be an important problem. A real 5. Murtha, J.A.: “Estimating Reserves and Success for a
understanding of the nature of this data is needed if accurate Prospect with Geologically Dependent Layers”, paper
assessments of future potential are to be achieved. It is SPE 30040 presented at the 1995 SPE Hydrocarbon
unfortunate that the field was developed so early that it Economics and Evaluation Symposium, Dallas, TX, Mar.
predated the industry’s understanding of the importance of 26-28.
accurate measurements. 6. Wang, B. and Hwan, R.R.: “Influence of Reservoir Drive
If the potential errors described above are added to the Mechanism on Uncertainties of Material Balance
normal field errors described in Fields 1 and 2, the error bars Calculations”, paper 38918 presented at the 1997 SPE
on the Field 3 production history would be large and would Annual Technical Conference and Exhibition, San
ultimately affect both the history match and the predictions of Antonio, Oct. 5-8.
future recovery. 7. Hwan, R.R.: “Improved Material Balance Calculations by
Coupling with a Statistics-Based History Matching
Conclusions Program”, paper SPE 26244 presented at the 1993
1. Roughly 80 billion tons (550 billion barrels) are at risk Petroleum Computer Conference, New Orleans, July 11-
due to historical data errors and uncertainties in fields that 14.
were developed without modern operating philosophies, 8. Murtha, J.A.: “Infill Drilling in the Clinton: Monte Carlo
tools and techniques. Techniques Applied to the Material Balance Equation”,
2. The problem is widespread in regions where most of the paper SPE 17068 presented at the 1987 SPE Eastern
expected remaining oil reserves are located, such as the Regional Meeting, Pittsburgh, PA, Oct. 21-23.
FSU, South America and the Middle East. 9. Murtha, J.A.: “Incorporating Historical Data in Monte
3. Historical data problems are more common than is Carlo Simulation”, paper SPE 26245 presented at the
normally assumed in normal field analysis. As more 1993 Petroleum Computer Conference, New Orleans, July
older fields are reviewed for rejuvenation potential, these 11-14.
problems will become more apparent and important. 10. Fair, W.B., “A Statistical Approach to Material Balance
4. A strong indicator of historical data problems is when it is Methods”, paper SPE 28629 presented at the 1994 SPE
necessary to make abnormal adjustments to the geologic Annual Technical Conference and Exhibition, New
model to explain reported production, injection, or Orleans, LA, Sept 25-28.
pressures. In these cases, the historical data may also 11. Spencer, J.A., and Morgan, D.T.K.: “Application of
need to be adjusted to achieve a reasonable match. Forecasting and Uncertainty Methods to Production”,
5. The older the field, the more likely it is that problems paper SPE 49092 presented at the 1998 SPE Annual
exist in the historical data. However, these problems are Technical Conference and Exhibition, New Orleans, LA,
not limited to old fields and can result when “old” Sept 27-30.
operating policies (e.g., poor measurement practices) are 12. Jensen, T.B.: “Estimation of Production Forecast
used. Uncertainty for a Mature Production License”, paper SPE
49091 presented at the 1998 SPE Annual Technical
Recommendations Conference and Exhibition, New Orleans, LA, Sept 27-
1. Each field has its own unique problems. The exploitation 30.
team needs to thoroughly understand the nature and
6 F.T. BLASKOVICH SPE 59471

13. Kaufman, R.L., Dashti, H., Kabir, C.S., Pedersen, J.M., Field”, paper SPE 37747 presented at the 1997 Middle
Moon, M.S., Quttainah, R. and Al-Wael, H.: East Oil Show and Conference, Manama, Bahrain, March
“Characterizing the Greater Burgan Field: Use of 15-18.
Geochemistry and Oil Fingerprinting”, paper SPE 37803 15. Craig, F.: The Reservoir Engineering Aspects of
presented at the 1997 Middle East Oil Show and Waterflooding, AIME, New York, N.Y. (1971).
Conference, Manama, Bahrain, March 15-18.
14. Pedersen, J.M., Moon, M.S. and Al-Ajeel, H. : “Data SI Metric Conversion Factors
Validation: Key to Development of an Integrated 1 metric Ton = 7 stock-tank barrels
Reservoir Model for the Wara Formation, Greater Burgan 1 m3 /ton = 5 scf/stb

Table 2 - Field 2 Production Database

W e l l s M M T o n s

C u m O i l P r o d u c e d b y 1 2 / 9 6 608 65.6
C u m W t r P r o d u c e d b y 1 2 / 9 6 608 79.1
C u m W t r I n j e c t e d b y 1 2 / 9 6 148 182.5
V o i d a g e R e p l a c e m e n t ( f r a c t i o n ) 1.17

Oil Producers converted to Injectors 125

W e l l s w i t h c o m i n g l e d p r o d u c t i o n o r i n j e c t i o n ? 20 ?

No Perforations, Production or Injection 248 ?

Table 1 - Remaining Proved World Oil Reserves
Perforations, but no Production or Injection Data 20 ?
Production, but no Perforations 27 1.5
REGION RESERVES (MMTons)
Injection, but No Perforations
Asia-Pacific 6,145 18 4.5
(16 injectors assumed
Western to have production perfs)
Europe 2,674
Eastern Europe & FSU 8,436
Middle East 96,235
Africa 10,777
North America 10,757
South America 12,785
WORLD TOTAL 147,810

Data from 1999 International Petroleum Encyclopedia

SPE 59471 HISTORICAL PROBLEMS WITH OLD FIELD REJUVENATION 7

160

140

120
Oil Reserves (billion metric tons)

100

OPEC
80 World
Old Fields

60

40

20

0
1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

Discovery Date (year)

Figure 1 - World Oil Reserves

6000

5000

4000
Oil Rate (Tons/day)

3000 Observed
Model

2000

1000

0
0 2 4 6 8 10

Time (years)

Figure 2 - Field 1 Oil Production Match

8 F.T. BLASKOVICH SPE 59471

1600

1400

1200
Water Rate (Tons/day)

1000

Observed
800 Model

600

400

200

0
0 2 4 6 8 10

Time (years)

Figure 3 - Field 1 Water Production Match

10000

9000

8000

7000
Injection Rate (Tons/day)

6000
Observed
Model
5000

4000

3000

2000

1000

0
0 2 4 6 8 10

Time (years)

Figure 4 - Field 1 Water Injection Match

SPE 59471 HISTORICAL PROBLEMS WITH OLD FIELD REJUVENATION 9

4%

3%
Importance

2%

1%

0%
0% 50% 100% 150% 200% 250%

Error

Figure 5 - Oil Production Well Errors

10%

8%

6%
Importance

4%

2%

0%
0% 5% 10% 15% 20% 25%

Error

Figure 6 - Water Injection Well Errors

 10

Reservoir Pressure (atms) Reservoir Pressure (atms)

M
ar

150
200
250
300
350
-8

200
250
300
350
400

3/1/87 7
Se
p-
87
9/1/87 M
ar
-8
8
3/1/88 Se
p-
88
9/1/88 M
ar
-8
9
3/1/89 Se
p-
89
9/1/89 M
ar
-9
0
3/1/90 Se
p-
90
9/1/90 M
ar
-9
1
3/1/91 Se
p-
91
9/1/91 M
ar
-9
2
3/1/92 Se
p-
92
M
9/1/92 ar
-9
3
Se
3/1/93 p-
93
F.T. BLASKOVICH

M
9/1/93 ar
-9
4
Se
3/1/94 p-
94
M
9/1/94 ar
-9
5
Se
3/1/95 p-

Figure 8 - Well 385 Pressure History

95
M
9/1/95 ar
-9
6
Se
3/1/96 p-
96
M
9/1/96 ar
-9
Figure 7 - Field 1 Average Reservoir Pressure History

7
Se
3/1/97 p-
97
M
ar
9/1/97 -9
8
3/1/98
SPE 59471
 SPE 59471

Oil Production (Tons/Month) Oil Production (tons/month)

M
ar M
-8
7

1000
2000
3000
4000
5000
6000

0
ar
-8

0
20000
40000
60000
80000
100000
120000
140000
160000
180000
200000

Se
p- Se 7
87 p-
87
M
ar M
-8 ar
8 -8
Se 8
p- Se
88 p-
88
M
ar M
-8 ar
9 -8
Se 9
p- Se
89 p-
89
M
ar M
-9 ar
0 -9
Se 0
p- Se
90 p-
90
M
ar M
-9 ar
1 -9
Se 1
p-
Se
91 p-
91
M M
ar
-9 ar
2 -9
Se 2
p-
Se
92 p-
92
M M
ar
-9 ar
3 -9
Se 3
p-
Se
93 p-
93
M M
ar
-9 ar
4 -9
Se 4
p-
Se
94 p-
94
M M
ar
-9 ar
5 -9
Se
5
Se
p-
95 p-
95
M M
Figure 9 - Field 1 Oil Production History

Figure 10 - Well 385 Production History

ar
-9 ar
6 -9
Se
6
HISTORICAL PROBLEMS WITH OLD FIELD REJUVENATION

Se
p-
96 p-
96
M M
ar
-9 ar
7 -9
Se
7
Se
p-
97 p-
97
M M
ar ar
-9
8 -9
8
11
 12

Water Injection (tons/month) Water Injection (Tons/Month)

M
ar
-8
M
7

5000

0
ar

10000
15000
20000
25000
30000
35000
40000
45000
50000
-8

0
50000
100000
150000
200000
250000
300000
350000
400000
450000

Se
p- Se 7
87 p-
87
M
ar M
-8 ar
8 -8
Se 8
p- Se
88 p-
88
M
ar M
-8 ar
9 -8
Se 9
p- Se
89 p-
89
M M
ar
-9 ar
0 -9
Se 0
p-
Se
90 p-
90
M M
ar
-9 ar
1 -9
Se 1
p-
Se
91 p-
91
M M
ar
-9 ar
2 -9
Se 2
p-
Se
92 p-
92
M M
ar
-9 ar
3 -9
Se 3
F.T. BLASKOVICH

p-
Se
93 p-
93
M M
ar
-9 ar
4 -9
Se
4
Se
p-
94 p-
94
M M
ar
-9 ar
5 -9
Se
5
Se
p-
95 p-
95
M M
ar
-9 ar
6
Figure 11 - Field 1 Water Injection History

-9
6

Figure 12 - Well 550 Water Injection History

Se Se
p-
96 p-
96
M M
ar
-9 ar
7 -9
Se
7
Se
p-
97 p-
97
M M
ar ar
-9
8 -9
8
SPE 59471
SPE 59471 HISTORICAL PROBLEMS WITH OLD FIELD REJUVENATION 13

25000

20000
Oil Production (MTons/year)

15000

History
Model

10000

5000

0
0 5 10 15 20 25

Time (years)

Figure 13 - Field 2 Oil Production Match

18000

16000

14000
Water Production (MTons/year)

12000

10000
History
Model
8000

6000

4000

2000

0
0 5 10 15 20 25

Time (years)

Figure 14 - Field 2 Water Production Match

14 F.T. BLASKOVICH SPE 59471

50000

45000

40000

35000
Water Injection (MTons/day)

30000

History
25000
Model

20000

15000

10000

5000

0
0 5 10 15 20 25

Time (years)

Figure 15 - Field 2 Water Injection Match

5.0%

4.0%

3.0%
Importance

2.0%

1.0%

0.0%
0 100 200 300 400 500 600 700

Error

Figure 16 - Field 2 Oil Production Error

SPE 59471 HISTORICAL PROBLEMS WITH OLD FIELD REJUVENATION 15

5%

4%

3%
Importance

2%

1%

0%
0 100 200 300 400 500 600 700

Error

Figure 17 - Field 2 Water Injection Error

1600 300

1400
250

1200
Oil or Water Rate (Tons/day)

200
Gas-Oil Ratio (m3/ton)

1000

800 150

600
100

400

50
200

0 0
26

29

32

35

38

41

44

47

50

53

56

59

62

65

68

71

74

77

80

83

86

89

92

95
19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

19

Time (years)
OIL WATER GOR

Figure 18 - Field 3 Production History