Choice of Development Concept—

Platform or Subsea Solution?

Petter Osmundsen, University of Stavanger

Summary normally have higher regularity over their producing life. New re-
A real choice exists today on a number of discoveries between plat- covery technology, which emerges after development has ended, is
form-based or subsea development solutions. Statistics from the often easier to adopt when a platform has been chosen.
Norwegian continental shelf (NCS) show that fields developed The recovery factor is defined as the proportion of the oil in a
with fixed platforms have a substantially higher recovery factor. reservoir that is recovered. A key concept in this context is stock-
The potential for a later commitment to improved oil recovery tank oil originally in place (STOOIP). “Stock tank” is the volume
(IOR) is determined largely by the original development solution. at normal pressure and temperature. STOOIP must not be confused
Through the use of cases and examples, this paper discusses prin- with oil reserves, which are the volume that can be recovered tech-
ciples for valuation of the enhanced flexibility offered by platform- nically and commercially (Osmundsen 2010). The recovery factor
based development solutions and sequential subsea solutions. It for offshore oil fields normally lies between 10 and 60%, but can
illustrates that valuing the various types of flexibility is difficult, reach close to 80% in certain favorable cases (Energy Information
which leads to the following question: Are development solutions Administration 2008).
being selected without taking sufficient account of option values? Approved oil-company plans at the end of 2010 would mean
that 54% of the oil in fields developed on the NCS remains un-
Introduction recovered (IOR Expert Committee 2010). Norway has achieved
Technological progress with subsea production has been rapid. high recovery factors compared with other countries. A global
Such installations can now be used in most conditions, and costs overview of recovery factors is provided in Sandrea and Sandrea
have been reduced sharply. A real choice exists today on a number (2007). They report an overall factor of 46% for the North Sea,
of discoveries between platform-based or subsea development so- and describe this as the highest in the world. According to Laher-
lutions. In particular, a subsea facility could be a good solution for rere (2006), the global average recovery factor is 27% (derived
fields with small resources or in deep water where the distance to from the Information Handling Services database, which covers
land or to existing platforms is short. The choice of concept is a approximately 11,500 fields). Nevertheless, substantial financial
complex business, with input from many interested parties and gains could be made from improving the recovery factor; an in-
technical disciplines. Examples of key developments on the NCS crease of just 1% in oil production beyond today’s approved plans
that faced a demanding choice of concept are Ormen Lange and could yield net revenues on the order of USD 20–30 billion at cur-
Snøhvit in the Norwegian Sea and Barents Sea, respectively. These rent oil prices (Melberg 2009). It is difficult to make accurate cost
fields have been developed with subsea solutions even though that estimates here, and it is consequently of equal interest to look at the
has required long tiebacks to land-based terminals. Platforms were corresponding gross revenue, which is on the order of USD 50–60
one alternative studied. billion. As always, revenues must agree with costs, but a poten-
Investment in subsea installations is lower, but drilling costs re- tial for profitability very probably exists for both government and
main high throughout the field’s producing life, and licensees may oil companies.
often have to pay tariffs to infrastructure owners. In other cases, The development concept is one element that influences the re-
the same partners own both the subsea field and the processing fa- covery factor, and that offers a choice. Reservoir, fluid, and rock
cilities, as with the aforementioned Ormen Lange and Snøhvit ex- properties are more important, but are determined by nature in the
amples. If, as in these cases, the development involves a tieback same way as porosity, permeability, and the quantity of gas dis-
of subsea facilities to a newly built land-based terminal, this will solved in the oil together with heavier components that can cause
be included as investment in the net-present-value (NPV) calcula- wax formation and raise oil viscosity, thereby hampering produc-
tions. On the other hand, when the choice is to tie back to an ex- tion. The recovery factor depends also on the efforts made by the
isting processing facility owned by the licensee, which could now oil companies to maintain production over time, including injection
or over time be used by other projects (owned by the same licensee of water, gas, and chemicals in addition to well workovers and new
or others), an opportunity cost must always be calculated for use drilling. But the choice of development concept has a great impact
of the capacity. Fixed platforms offer a number of advantages, on the cost of subsequent IOR work. Therefore, it is interesting for
which need to have a value put on them. Such installations permit government and companies to study the validity of decision criteria
a flexible drainage strategy, particularly if the platform has its own for concept choice—the extent to which these take account of the
drilling facilities. They offer lower marginal costs for IOR cam- relationship between concept choice and recovery factor.
paigns after a few years of learning lessons on the field, and they
Real Options in Oil Recovery
The potential for a later commitment to IOR is determined to a
great extent by the original development solution. One based on
Copyright © 2013 Society of Petroleum Engineers a dedicated drilling rig, for instance, will normally have greater
Original SPE manuscript received for review 28 February 2012. Revised manuscript
potential than platforms without such facilities or than subsea so-
received for review 29 August 2012. Paper (SPE 163144) peer approved 7 January 2013. lutions in which a mobile rig must be chartered each time. This

64 Oil and Gas Facilities  •   October 2013

 Recovery Rate Percentage Deviation in Recovery Rate
60 180
160
Recovery Rate, %

50
140
40 120

Percent
30 100
Facility
80 Percentage deviation
20 Subsea platform vs. subsea
60
10 40
0 20
1990 1995 2000 2005 2010 0
Year 1990 1995 2000 2005 2010
Year
Fig. 1—Average recovery factors for fields with a platform and Fig. 2—Percentage difference in average recovery for fields with
those developed with subsea wells. Platforms are defined here fixed platforms and those developed with subsea completions
as fixed structures with a drilling module (Nordvik et al. 2010). (Nordvik et al. 2010).

affects not only the flexibility for, but also the marginal cost of, facilities have a substantially higher recovery factor. This is illus-
workovers or new wells. trated in Fig. 1.
One advantage of subsea installations is lower initial invest- The difference in recovery factor between fields with fixed plat-
ment. On the other hand, costs are higher for operation and mainte- forms and those developed with subsea completions equals seven
nance, tariffs may often have to be paid for processing, flexibility percentage points. For fields included in the statistics, this trans-
is lost, and it is far more expensive to drill new wells or imple- lates into 17% higher production on average with a platform. The
ment necessary changes to existing ones. Installing a platform with reason for the difference is that, while the recovery factor is calcu-
drilling facilities makes it easier and less expensive to intervene in lated in relation to the STOOIP, the production increase is calcu-
wells, run measuring devices, and identify and diagnose improve- lated in relation to existing output; in other words, the denominator
ment possibilities. Opportunities for injection are greater, and more in the latter fraction is substantially smaller. We can see from
wells can be drilled. It is also simpler and less expensive to im- Fig. 2 that the percentage difference fell sharply until 1998
plement necessary changes, including alterations to the drainage (when it was 13%), and thereafter flattened out, although with
strategy. An improvement measure on a subsea well often requires some fluctuations.
five times the earnings potential than would be needed for an in- When interpreting these data, some caveats are worth pointing
tervention in a platform well. Delays to well intervention are one out. When using statistics, the possibility of sampling errors must
consequence of this. The backlog in well maintenance has led to always be borne in mind. Ideally, the recovery factor for different
production losses that cannot be recovered and to the downgrading development concepts should be compared for the same field;
of reserves (IOR Expert Committee 2010). At the same time, a plat- however, that is not possible. Thus, parts of the discrepancy in the
form solution will provide greater assurance that the position has average extraction rate may be caused by factors other than devel-
been understood while providing a better database and lower op- opment concept. One such factor is reservoir size. Because smaller
erational risk, which relates in part to weather conditions (drilling fields are not able to carry the higher capital investment of a plat-
from a platform or a jackup rig cantilevered over a wellhead instal- form development, the average reservoir size is lower for subsea
lation is seldom halted by bad weather). A platform solution avoids fields. Smaller fields, in general, have lower extraction rates than
the restrictions on well numbers imposed by a subsea development. larger fields (e.g., because some IOR techniques are not profitable
Operations can also be optimized regardless of sharply fluctuating on a small reservoir and because larger fields can prolong the ex-
rig rates. traction time by becoming hosts for tie-ins for other reservoirs).
On the other hand, there are other operational risks that the plat- Developments proceed with incomplete information, but the com-
form with rig suffers from and the subsea alternative does not [e.g., panies know a good deal from interpreting seismic and well data.
the simultaneous operations of production and drilling (or comple- Because they are often able to make a concept choice suited to the
tion or workover) on the same platform]. Platform solutions also reservoir, the variation in recovery factor between platforms and
have their restrictions. The number of wells from a single point subsea completions (as shown in Figs. 1 and 2) may be somewhat
(the case of a platform-with-a-rig solution) is restricted not only be- exaggerated. Moreover, the data series does not cover technolog-
cause of the number of slots in the template, but also because direc- ical developments over the last few years that are likely to have
tional/horizontal drilling is feasible for reservoir targets up to only reduced the benefit of a platform development. Examples are light
approximately 7 km. For the subsea solution, restrictions apply for well intervention, multilateral wells, subsea processing, downhole
subsea wells producing to existing fixed platforms, but it is pos- measurements and flow control, and improved reservoir moni-
sible to connect subsea wells to floating production units (e.g., toring. In comparison of platform (with rig) vs. subsea solutions,
semisubmersibles without rigs, and floating production, storage, it should also be noted that in the 1990s, the oil price was low and,
and offloading units). The Marlim oil field in Brazil, for example, in many cases, the subsea alternative could be the only one with
has more than 100 subsea wells connected to several floating pro- positive NPV. Similarly, some discoveries were fields with small
duction units without rigs. volumes so that, again, the subsea concept could be the only eco-
The threshold for making changes to subsea wells is often very nomic solution. Of course, insisting on a platform concept in these
high. It is possible, for instance, to find oneself in conditions in cases would not improve the recovery factor, because the fields in
which rig rates are increased for many days because of bad weather. these cases would not have been developed. Still, there are many
Platform wells also have better production regularity, while me- field developments in which both subsea and platform concepts are
chanical damage can, as a rule, be repaired and wells brought back possible from an economic point of view. In these cases, investing
on stream in reasonable time. Taken together, these considerations in a platform concept that offers more flexibility may pay off be-
mean that developments based on platforms with their own drilling cause of the uncertainties associated with reservoir engineering.

October 2013  •   Oil and Gas Facilities 65

In particular, this applies to complex reservoirs, for which more method demands well reconfiguration (e.g., tubing replacement).
data acquisition, sidetracks, and well interventions are required to These less-expensive exercise costs for workover, recompletion,
achieve high recovery rates. Hence, valuation of production options and IOR options (and the increased availability of rigs) generate a
is important. higher production and oil recovery for a platform-with-rig alterna-
Real-option theory is a well-developed discipline that makes it tive than with subsea-based development.
possible to price a number of real options. A key textbook in this To ensure that all real-option effects related to concept choices
area is Dixit and Pindyck (1994). In my experience, however, ex- are included, it could make sense to use simpler models, such as
isting oil-company models fail to pick up all real-option elements. sensitivity analyses, which take into account the differing drilling
More sources of flexibility exist than those shown in Table 1, in- costs and production volumes related to the various options.
cluding the concept that subsea solutions require developments in A simple approach to the issue of development with a platform
rig rates to be modeled. Conditions could also arise in which pro- or with a subsea solution is to regard this as a classic choice be-
duction is lost because of rig shortages. This could be caused by a tween expenditure today vs. expenditure tomorrow. A platform-
lack of tradition or by difficulties associated with the valuation of based development involves a higher initial investment, but lower
these options. drilling costs and tariff savings over the field’s producing life.
Closed-form option models are not able to analyze many of the However, the difference in cost structure has an additional effect,
real options listed in petroleum projects. This is partly because the which represents the main point of this article—lower post-devel-
latter are complex, partly because they are not independent, and opment drilling costs also yield a higher recovery factor and, there-
partly because the option models, which originate in the pricing of fore, increased revenues. In the following section, I will review a
securities, build on assumptions that are inappropriate for choosing simple example that can illustrate the effect on the income side.
concepts in petroleum developments (Pilipovic 2007).
Table 2 illustrates points at which economic conditions for pe- Example
troleum projects in general deviate from those of stock options, The financial effect of increased production on the choice of a plat-
making it more difficult to price petroleum options. Real-option- form-based development will depend critically on whether the ex-
pricing theory has developed further and, currently, is able to ac- pected increase in volume takes the form of higher ongoing output
commodate several of those pricing challenges adequately (Haug (greater plateau production) or an extended producing life for the
1997). However, for the particular case of pricing production flex- field. The first of these effects could be obtained when a develop-
ibility, additional challenges exist. Concepts such as reservoir ment is tailored optimally to the reservoir. Succeeding in that (with
complexity and technical flexibility are not readily captured by the aid of good reservoir understanding and a reservoir that is not
economic option models, because we do not have knowledge of too complex) means a high recovery factor can also be achieved
relevant distribution functions, nor are these functions likely to sat- with a subsea solution. If, on the other hand, the reservoir is com-
isfy regularity assumptions on which pricing formulas are based. plex and surprises are encountered, the increased flexibility of-
Still, it is possible to capture parts of the real-option values de- fered by a platform will provide higher plateau production. In other
scribed in Table 1: higher value for the option to expand the pro- cases, the greater flexibility will be experienced primarily in the
duction (IOR option) as a result of the lower exercise price to field’s final phase by allowing its producing life to be extended. Be-
implement the IOR project. The additional real-option value of the cause of discounting, volume increases in the final phase will exert
platform with rig compared with the subsea alternative is caused less influence on the NPV.
mainly by the lower exercise price (or exercise cost) of options re- These effects can be illustrated by a simple calculation. I am as-
lated to wells as a result of the less-expensive rig in platform-based suming here a model field that can produce 100, 150, or 200 mil-
development compared with rigs in floating units. This lower ex- lion bbl of oil from a platform-based development. By applying
ercise price, together with the higher rig availability, increases the the average recovery factor for the NCS in 2008 (47% for plat-
option value of (a) workover in wells that stopped production be- forms and 40% for subsea completions), it is determined that the
cause of technical problems or workover in wells aiming to im- corresponding recovery for a subsea solution would be 85, 127, or
prove production with acid fracturing or another method. The high 170 million bbl. A lead time of 3 years is assumed. For simplici-
rig rate makes the exercise price too high to perform some work- ty’s sake, the production rise from choosing a platform rather than
overs in the subsea alternative; (b) recompletion opportunities (this a subsea solution is assumed to occur on a straight-line basis over
is important for fields with different reservoirs at different depths 15 years when the increase takes effect in plateau output. When
or many noncommunicating layers). The lower exercise price for the improvement alternatively comes at the end of the field’s pro-
platforms with rigs makes it more likely to exploit these reinvest- ducing life, it is assumed to be allocated on a straight-line basis
ment opportunities than in the case of subsea development; and (c) over 5 years, so that the overall production period extends to 20
IOR investments in the end of oilfield life, mainly when the IOR years. The real discount rate is set at 10% (Boston Consulting

TABLE 1—REAL OPTIONS IN THE CHOICE TABLE 2—COMPARISON OF A STANDARD

OF CONCEPT FOR OFFSHORE PETROLEUM STOCK OPTION AND A COMPLEX ENERGY OPTION
DEVELOPMENT—INCREASED OPPORTUNITIES
FROM CHOOSING A PLATFORM
Standard Stock Option Petroleum Projects
Real Options Related to Platform-Based Developments
Price option Option on price, volume,
and timing
Price process: random walk
Flexible drainage strategy Price process: mean reversion
Independent periods
Technical flexibility, greater potential Not independent
Good liquidity
Financial flexibility, lower marginal costs for extra measures Lack of liquidity
Arbitrage
Lower operational risk Limited arbitrage
Absence of seasonal effects
Greater regularity Seasonal effects in gas markets
Absence of convenience yield
Considerable convenience yield

66 Oil and Gas Facilities  •   October 2013

Group 2005), oil price at USD 90/bbl in real terms, and the US Revenue of Subsea vs. Platform
dollar exchange rate at NOK 6. 1,200
From Fig. 3, we see that the gain in NPV of revenue through
1,000

NPV, USD million

the rise in volume could be as high as USD 1 billion. This rev-
enue increase is supplemented by the NPV of savings in operating 800
costs over the field’s lifetime from a platform-based develop- NPV increased plateau
600
ment solution, which includes lower drilling costs and tariffs paid NPV extended lifetime
to infrastructure owners over the field’s producing life. The dis- 400
counted sum of these two effects—higher revenues and saved
200
operating costs—represents the rise in initial investment one
should be willing to bear to opt for a platform-based solution. 0
Fig. 3 shows that this willingness to pay varies substantially with 100 150 200
expected reserves. Million bbl
This is only a rough example. Other assumptions could obvi- Fig. 3—The rise in revenue for a model field measured by NPV
ously yield different results. A lower rate of return would boost in USD millions for a platform-based vs. a subsea develop-
NPV. Interest rates have fallen substantially since 2005, and are not ment, with total production from the model field of 100, 150, and
expected to rise in the near future, which would encourage lower 200 million bbl.
required rates of return. The same effect would be achieved by as-
suming a real rise in oil prices in the time to come. A different pro- tion from subsea wells—looks to be a more-convenient method to
duction profile, which takes longer to reach plateau, would reduce develop petroleum fields using modern real-option concepts.
the NPV to a certain extent. When a development decision is taken, knowledge of the field
The difference in recovery factor between subsea solution and will be limited, including future opportunities and challenges that
platform-based solution is the most important parameter here; it is might arise in its producing life. Flexibility is crucial to a valuation.
also the most difficult to estimate. By using average figures, I im- The danger is that the greatest weight will be given to initial invest-
plicitly assume an arbitrary decision, which is probably incorrect. ment savings, because these are the easiest to tackle or because a
If the oil companies succeed systematically in making a concept short-term approach is being taken. The development team will be
choice tailored to the reservoir, the expected difference between satisfied if it can achieve a reasonable project, and the company and
the two development concepts will be lower than average figures its present management receive positive media coverage. However,
for the NCS suggest. what matters in the long run for an oil company is the life-cycle
Subsea solutions are often selected because a platform-based economics expressed in the project’s NPV, including the relevant
development would not be profitable—initial investment is sig- options available. However, it must be stressed in this context that
nificantly lower with seabed installations. In deep water, a subsea realizing these options could involve substantial additional costs
approach is often the only one possible. However, the appropriate that would need to be taken into account.
solution for many developments is a matter of doubt. Large reserves Oil companies have developed financial models that take into
point toward a platform-based concept because achieving a high re- account many such options. Accuracy in applying these models de-
covery factor makes good economic sense. Additionally, a complex pends on good communication between the various disciplines and
reservoir (increasingly common on the NCS) favors a platform ap- close collaboration. Decision-support models have been improved
proach because such reservoirs require greater flexibility. In such substantially, but they call for suitable input parameters. According
a case, a subsea facility would mean high costs in the form of new to company financial teams, they do not always get these from their
wells and workovers, and major assets could remain in the ground petroleum-technology colleagues when seeking to calculate real-
because the wrong development solution was chosen. On the other option values. The option models are often complex and difficult
hand, a platform could also represent an erroneous approach if the to solve, and could have limited freedom in terms of input format.
reservoir has been overvalued. The resulting development could Obtaining suitable input depends on detailed knowledge of the de-
fail to justify the investment cost. One could argue that complex cision-support models among petroleum technologists and on their
reservoirs demand the exercise of learning options (investment in willingness to estimate suitable parameters. Ideally, the various
information) before committing huge investment in the overall oil- sides should also agree on what constitutes suitable input to the
field development. Complex reservoirs typically have relevant dy- decision analysis. If the input parameters are not tailored to the
namic uncertainties (e.g., stabilized productivity index after some models, the danger is that the size of the initial investment dom-
weeks or months of production, which can be very different from inates when decisions are made. Naturally, developing decision
the initial productivity index, mainly in carbonates) that can be re- models tailored to available and relevant parameters also poses a
duced with one or a few subsea wells. The platform-with-rig so- challenge to the economists. A problem the latter face is that time
lution is best suited for oil fields with low uncertainty so that the will be a critical factor. The financial analysis is the final link in
location of the template and the processing capacity/top-facilities the chain, and the analysts have little time available. This does not
decisions can be made with a reasonable degree of certainty. Note seem to be the best point for the oil companies to reduce the time
that from the platform template location, it is not feasible (techni- taken—quite the contrary, in fact.
cally and/or economically) to reach reservoir targets with a distance Defending more-expensive solutions on the basis of gut feeling
of more than approximately 7 km from the template location with and industrial instinct calls for considerable courage on the part of
directional or horizontal wells. Therefore, the definition of the plat- management. It is frequently the case that quantitative effects dom-
form template location is important, and, for complex fields, could inate qualitative ones—the former are often harder to challenge and
be important to the dynamic information (from initial production) easier to audit afterwards. An increased concentration on auditing
provided by the subsea alternative, at least for the first wells. and transparency can have the unintended consequence that exces-
In short, there are real options that favor the platform-with-rig sive weight is given to easily measureable conditions when making
solution (workover, recompletion, and IOR; the last two are options decisions. At certain times, too, management of the operator com-
to expand the production), and real options that favor the subsea so- pany (or the partners in the license) works with a self-imposed ra-
lution if applied sequentially (learning options). In many cases, the tioning of capital, and may then opt for the less-expensive solution,
combination of these two solutions—a few subsea wells in the be- even though this yields a lower expected NPV (Osmundsen et al.
ginning followed by an optimized platform by use of the informa- 2006; 2007).

October 2013  •   Oil and Gas Facilities 67

Supplementary Considerations. It was demonstrated in the pre- a well intervention in the light of high rig rates. If this condition
ceding example that platform-based developments provide greater remains fairly constant for a few years, the realization with hind-
flexibility, which permits a higher recovery factor and, thereby, sight is often that one should have intervened earlier and made
substantial additional revenues. However, a number of advantages more money, but that it is now definitely too late. Additionally,
of subsea systems have not been taken into account in this discus- rig avai ability and total drilling costs were not given enough em-
sion and numerical example. An important characteristic of subsea phasis to ensure optimum earnings. The combination of small re-
solutions is that they simplify a phased delineation and develop- serves and an uncertain upside for remaining resources in a field
ment of fields, and, therefore, normally provide an earlier start to led, and continues to lead, to well intervention on subsea installa-
production with the gathering of useful information. They usu- tions being neglected.
ally involve predrilling, so that plateau production is reached more Plans to build light intervention rigs existed as early as the
quickly. Predrilling can also be conducted with fixed installations, late 1980s, but foundered through lack of collaboration in the in-
but that involves additional investment and risk. Faster develop- dustry, new business models in the oil companies, and uncertain
ment and shorter time to plateau almost always increases NPVs. crude prices in the early 1990s. The fields were in full plateau pro-
However, this argument assumes that a rig is available. To the ex- duction, and nobody wanted to become unpopular by proposing
tent that a tie-in is required to an existing installation, this opportu- that a great amount of money be spent on something that was
nity must be available with the desired capacity at the anticipated a problem for the future. The concentration on short-term produc-
time. This does not apply if subsea wells are linked to floating pro- tion indicators could have played a part here. Well tools were devel-
duction units. These units may have valuable options to abandon oped circa 1990, when the subsea licensees joined forces to create a
because they can be relocated to another petroleum field after petro- pool of installation and maintenance equipment. The same should-
leum-field exhaustion. Experience shows that these requirements have been done for light well-intervention vessels. As illustrated
are not always met. It was necessary to wait for spare capacity until in Fig. 2, developments have shown that this was an erroneous de-
other fields went off plateau, and tariff negotiations took time. A cision. The well-maintenance backlog is now substantial and has
tieback may also require modifications, which have often turned led to production losses that cannot be retrieved (consider the
out to be more expensive and more time-consuming than the NPV downgrading of reserves on the Halten Bank).
calculations assumed. However, it is clear that not having to de-
sign, order, and build one or more platforms with equipment and Case: Gullfaks South
so forth helps in terms of timing. Pumps, compressors, and tur- Gullfaks South lies due south of Gullfaks in the northern North Sea.
bines/generators have all taken several years to deliver in periods. It has been developed with 12 subsea templates tied back to
I have been unable to obtain figures on development times for al- the Gullfaks A and C platforms.
ternative concepts.
The number of well slots on a platform is determined before Description of the Field.
construction begins. Extra wells must wait for spare slots (addi- • Discovery year: 1978
tional slots are inexpensive if they are included from the start). Ad- • Development approved: 29 March 1996
ditional slots may therefore pose a greater challenge on a platform • On stream: 10 October 1998
than with a subsea solution to which more templates can be in- • Operator: Statoil Petroleum A/S
stalled. The challenge is to secure enough capacity in pipelines and • Present licensees: Petoro A/S 30.00%, Statoil Petroleum
control systems. Pre-investment is cheaper than wisdom after the A/S 70.00%
event, but has an immediate impact on NPV calculations. Another
strength of subsea solutions is that drilling locations can be dis- Gullfaks South has been developed in two phases. The plan for
persed to optimum points in relation to the reservoir, avoiding un- development and operation (PDO) of Phase I embraced the produc-
necessarily long and expensive wells. tion of oil and condensate from the 34/10-2 Gullfaks South, 34/10-17
Payment for tie-in and tariffs for subsea solutions primarily in- Rimfaks, and 34/10-37 Gullveig deposits. Approved on 8 June 1998,
volve marginal costs on the platforms and a share of the fixed op- the PDO of Phase II embraced the Brent group in Gullfaks South.
erating costs. Should a new platform be built, all operating costs The 34/10-47 Gulltopp discovery was incorporated in Gullfaks South
must be borne by the discovery itself. However, this difference is during 2004. Gulltopp was produced through an extended-reach well
only relevant for a tieback phased in toward the end of a field’s drilled from Gullfaks A. The PDO for Rimfaks IOR and the 33/12-8
producing life; all costs must otherwise be met by the new fields. A Skinfaks discovery was approved on 11 February 2005, and em-
Major unexpected maintenance-related costs have arisen for fields braced a new template and a satellitewell. Incorporated in Gullfaks
in their final phase. As a rule, all tied-in fields must contribute to South, Skinfaks came on stream in January 2007.
meeting these, and a subsea solution can quickly prove to have The Gullfaks South reservoirs lie in Brent group sandstones from
been suboptimal in such circumstances. the middle Jurassic, and in the Cook, Statfjord, and Lunde formations
of the early Jurassic and late Triassic. Production occurs from the
Inadequate Well Maintenance Brent group and Statfjord formation. These reservoirs lie 2400–3400
The main problem facing subsea developments is that the m deep in rotated fault blocks. Gullfaks South’s reservoirs are seg-
threshold for new infill wells and well interventions is too high. mented extensively by many faults, and the Statfjord formation has
Active efforts are being made by the industry to lower this through poor flow properties. The other formations have fairly good reser-
the use of cheaper rigs, light well-intervention vessels, and stan- voir quality.
dardized solutions. Production from Gullfaks South is now being pursued by pres-
Bente Nyland, director-general of the Norwegian Petroleum Di- sure reduction after gas injection ceased in 2009. On Rimfaks, the
rectorate, has said that the maintenance backlog for subsea systems Brent group is producing with full pressure maintenance by gas in-
represents a challenge in the work of recovering the profitable re- jection, while the Statfjord formation has partial pressure support by
serves from existing fields on the NCS. “Many wells are out of the same means. The Gullveig and Gulltopp deposits are being pro-
operation,” she told Offshore.no. “Subsea developments present duced by pressure reduction and natural waterdrive, and their output
many advantages, but some challenges as well. And the industry will be influenced by Gullfaks production. Oil is piped to Gullfaks A
must put better maintenance systems in place” (Stangeland 2011). for processing, storage, and export by shuttle tanker, while the rich
So why have subsea wells not been maintained? Reserves fre- gas is processed on Gullfaks C and exported by means of Statpipe to
quently represent a conservative figure, and such estimates may Kårstø for further processing and dry-gas export to continental Eu-
often indicate in a given year that too little oil remains to justify rope (Nordvik et al. 2010).

68 Oil and Gas Facilities  •   October 2013

Controversial Development Solution. Gullfaks South is an ex- lation is being considered for the Luva field in 1270 m of water,
ample of a controversial choice between a platform and a subsea in- for instance.
stallation. The project was regarded as marginal, and earlier devel- It is also a good choice for small fields and reservoirs with a low
opments on Gullfaks (all platform-based) had involved high capital level of complexity. The technological progress made in coopera-
spending (with substantial overruns) and are viewed with hindsight tion with the major suppliers, a number of whom have their main
as having low profitability. Gullfaks South lies in relatively shallow base in Norway, has been useful and necessary, and has represented
water, and the reservoir was known to be complex. An optimistic an impressive export success. Continuous advances in subsea tech-
plan was drawn up with a minimum of wells. Originally, discus- nology have also gone some way in reducing the disadvantages of
sions on the choice of solution indicated that a platform could be subsea developments. When choosing a concept, it must be con-
defended if recovery were increased by 4 or 5%. Disagreement pre- sidered that topside technology developments and new-production
vailed in the licence over the development solution, but the major- solutions devised after the development date will often be easier
ity was convinced that it would be possible to achieve a recovery to adopt if a platform has been chosen. Pilot projects are essen-
factor similar to those of the other Gullfaks fields, even with a sub- tial for assessing alternative IOR methods, both present and future.
sea installation and despite the complex reservoir. Gullfaks South’s These are easier to pursue from a fixed installation; therefore, plat-
wells were drilled by a semisubmersible. Progress was poor, and form-based developments are favorable for continued innovation
costs were doubled. While platform-based developments also ex- on the NCS.
perience cost overruns (such as those on Gullfaks), these are of a This analysis has illustrated that the choice of concepts is com-
much lower order of magnitude (in percentage terms). A number of plex, with input from many parties and technical disciplines. Es-
problems have been experienced during the production phase that tablishing good communication is crucial here. When choosing a
could have been resolved better with a platform solution. According concept, it is often impossible to establish which solution is un-
to unofficial estimates, 20–25% of the reserves will be recovered ambiguously and objectively the best because so many sources of
compared with 60–70% for the other Gullfaks fields. The loss of uncertainty exist. In such circumstances, decisions are influenced
reserves is substantial, and, even allowing for possibly greater res- not only by knowledge but also by power. The relative strengths of
ervoir complexity, industry observers maintain that Gullfaks South the various technical disciplines (reservoir, drilling, facilities, and
could probably have attained a recovery factor of approximately project execution) will mean a great deal in practice. This is diffi-
40% with a fixed installation and a drilling rig constantly available. cult to handle in all organizations. Much can be achieved through
The lessons have hopefully been learned from this experience. the requirements and internal control bodies established by the
We see that many NCS developments have opted for a platform, in- company for work processes and the way in which assignments
cluding Ringhorne, Kvitebjørn, Gudrun, and Valemon. Relatively should be handled.
high oil prices at the decision point may have been a factor here. In addition, it is important that a balance of power exist between
these disciplines. The limited power and influence of people with
Conclusion subsurface expertise represents a problem in this context. There are
Developers have eventually become better at and more conscious several reasons for this. In numerical terms, the petroleum-tech-
about implementing real options in their decision-support systems nology disciplines (including geology, geophysics, reservoir engi-
when choosing development concepts for petroleum fields. There neering, and production engineering) form a relatively small group.
are real options favoring the platform-with-rigs solution (work- Furthermore, a culture of seeking senior executive positions no
over, recompletion, and IOR; the last two are options to expand longer seems to exist within Norway’s petroleum-technology dis-
the production), and real options favoring the subsea solution if it ciplines, as it does among economists and in part of the facilities
is applied sequentially (learning options). In many cases, the com- discipline. Efforts should be made to correct this imbalance, partly
bination of these two solutions (a few subsea wells in the begin- by adjusting the composition of company managements and partly
ning followed by an optimized platform by use of the information by taking more care to include arguments from petroleum technolo-
from subsea wells) looks to be a more-convenient approach to de- gists in decision processes.
velop the petroleum fields using modern real-option concepts. When the subsurface community comes up with a new idea, it
But, are developers taking into account all of the relevant op- is met with a well-nourished structure of control that consists not
tions? In practice, the position is probably that the large number of hunters, but of controllers and critics. These functions are also
of complex and mutually dependent real options available in important, but a balance must exist. Furthermore, the facilities dis-
such circumstances does not fit completely with existing deci- cipline can have its own agendas that do not always coincide with
sion models. Model calculations must be supplemented accord- high reservoir utilization. Subsurface expertise needs support and
ingly by judgements. It is important that petroleum-technology backing in executive management. This should be perceived as nat-
expertise be incorporated in such decisions. This case, perhaps, ural, because the greatest challenges to the oil companies for the
also represents an example of the way in which decision makers moment are on the resource side, related to production curves and
can be influenced trongly in certain circumstances by “the latest reserves replacement. It is appropriate therefore, that subsurface
experience,” and that their perspective can thereby become sub- expertise strengthen its position in the top management of compa-
optimal. At certain times, the perspective at the decision point nies, through the creation of a post of resource vice president, for
seems to be the lowest possible initial investment. Accordingly, it is instance. The top management and board should have a cross-dis-
important that companies work systematically on learning and ex- ciplinary composition, and a number of considerations indicate that
perience transfer in a decision-making context. subsurface expertise is not represented adequately.
Another relevant question is whether the basic estimates used as
input to the decision models are the best. Experience from the Acknowledgments
NCS and the UK continental shelf shows that the number of wells re- The author wishes to extend his appreciation to a number of spe-
quired in a field development is often underestimated—by 30%, ac- cialists in the petroleum administration and the oil sector for helpful
cording to an unofficial estimate. This points to a platform based suggestions and comments. The author also wishes to express his
solution in which drilling is less expensive once the initial invest- gratitude for constructive comments in connection with presenta-
ment has been made. If real options and the best cost estimates are tions at the seminar on IOR held by the Norwegian Petroleum So-
not taken adequately into consideration in the decision analysis, a ciety on 8 June 2011; the Norwegian Technical Science Academy
substantial IOR potential could have been lost as early as the choice on 16 February 2011; the Norwegian Petroleum Directorate on 30
of development solution. A subsea facility is often a relevant option September 2010; the Norwegian IOR expert committee on 15 April
in very deep water. Some exceptions exist here; a floating instal- 2010; the Department of Petroleum Engineering at the Univer-

October 2013  •   Oil and Gas Facilities 69

sity of Stavanger on 11 May 2010; and the 2010 Petrosam confer- Petroleum Directorate, Stavanger, Norway (June 2010), http://
ence on 9 June 2010. The Research Council of Norway is thanked www.npd.no/en/Publications/Facts/Facts-2010/.
for financing. Osmundsen, P. 2010. Chasing Reserves – Incentives and Owner-
ship. In Energy, Natural Resources and Environmental Eco-
References nomics, E. Bjørndal, M. Bjørndal, P.M. Pardalos, and M.
BCG. 2005. Investment Criteria, Methods, Decision-making Cul- Rönnqvist, 19–39. Berlin: Springer-Verlag.
tures Benchmarking—May 2005 Results. Report, The Boston Osmundsen, P., Asche, F., Misund, B. et al. 2006. Valuation of In-
Consulting Group, Boston, Massachusetts (October 2005). ternational Oil Companies. Energy Journal 27 (3): 49–64.
Dixit, A.K. and Pindyck, R.S. 1994. Investment Under Uncertainty. http://dx.doi.org/10.5547/ISSN0195-6574-EJ-Vol27-No3-4.
Princeton, New Jersey: Princeton University Press. Osmundsen, P., Mohn, K., Misund, B. et al. 2007. Is oil supply
Energy Information Administration (EIA). 2008. Defining the choked by financial market pressures? Energy Policy 35
Limits of Oil Production. International Energy Outlook 2008, (1): 467–474. http://dx.doi.org/http://dx.doi.org/10.1016/j.
Annual Report No. DOE/EIA-0484(2008), EIA/US DOE, enpol.2005.12.010.
Washington, DC (September 2008), http://www.eia.doe.gov/ Pilipovic, D. 2007. Energy Risk: Valuing and Managing Energy
oiaf/ieo/oilproduction.html. Derivatives, second edition. New York: McGraw-Hill.
Haug, E.G. 1997. The Complete Guide to Option Pricing For- Sandrea, I. and Sandrea, R. 2007. Global Oil Reserves-1: Recovery
mulas. New York: McGraw-Hill. factors leave vast target for EOR technologies. Oil Gas J. 105
IOR Expert Committee. 2010. Økt utvinning på norsk kontinen- (41): 44–48.
talsokkel: En rapport fra utvinningsutvalget (Increased pro- Stangeland, G. 2011. Subsea-brønner står uvirksomme. Offshore.
duction on the Norwegian continental shelf: A report from the no - Norway (14 January 2011), http://www.offshore.no/ny-
recovery selection (in Norwegian). Report, Publikasjonskode heter/sak.aspx?id=31552.
Y-0115 B, OLJE- OG Energidepartementet, Oslo, Norway (22
September 2010), http://www.regjeringen.no/upload/OED/
pdf%20filer/Oktutvinning.pdf.
Laherrere, J. 2006. Oil and gas: what future? Presented at the Gron-
ingen Annual Energy Convention, Groening, The Nether- Petter Osmundsen is a professor of petroleum economics at the Univer-
lands, 21 November. sity of Stavanger, Department of Industrial Economics and Risk Man-
Melberg, E.V. 2009. No prizes for lack of boldness. Norwegian agement. In 1992–93, he was a research fellow at the Massachusetts
Continental Shelf 2 (2009): 4-6, http://www.npd.no/Global/ Institute of Technology and Harvard University. Many of Osmundsen’s
Engelsk/3-Publications/Norwegian-Continental-Shelf/PDF/ research topics involve both economic and technical issues, and he
Norwegian-Continental-Shelf-No-2-2009(pdf).pdf. has published approximately 50 articles in international journals of
Nordvik, F.M., Verlo, S.B., and Zenker, E. ed. 2010. Facts 2010: economics, business management, and petroleum engineering. Os-
The Norwegian Petroleum Sector. Annual Report, Norwe- mundsen holds a PhD degree in economics from the Norwegian School
gian Ministry of Petroleum and Energy and the Norwegian of Economics and Business Administration in Bergen.

70 Oil and Gas Facilities  •   October 2013