JPT 2017-Marzo
JPT 2017-Marzo
JPT 2017-Marzo
AquaWatcher
WATER ANALYSIS SENSOR
An Official Publication of the Society of Petroleum Engineers. Printed in US. Copyright 2017, Society of Petroleum Engineers.
TECHNOLOGY FOCUS
We have the
45 HYDRAULIC FRACTURING
Zillur Rahim, SPE, Senior Petroleum Engineering Consultant,
superpower
Saudi Aramco
70 SEISMIC APPLICATIONS
Mark S. Egan, SPE, Consulting Geophysicist
VISURAY ION
X-RAY VIS
The complete SPE technical papers featured in this issue are available
free to SPE members for two months at www.spe.org/jpt.
RELIABLE, REAL-TIME DATA
AT 200°C
Breaking boundaries to deliver the difference.
When our client told us that they needed to collect high-quality logging-
while-drilling data at circulating temperatures up to 200°C, we knew it
wouldn’t be easy. But we also knew that through hard work, cooperation,
and ingenuity, we could make it happen.
26
Matthias Meister, Baker Hughes
2016 President
Nathan Meehan, Baker Hughes SOUTH ASIA AND THE PACIFIC
Salis Aprilian, PT Badak NGL
2018 President
SOUTHWESTERN NORTH AMERICA
Darcy Spady, Broadview Energy
Libby Einhorn, Concho Oil & Gas
Vice President Finance
WESTERN NORTH AMERICA
Roland Moreau, ExxonMobil Annuitant
REGIONAL DIRECTORS
Andrei Popa, Chevron
TECHNICAL DIRECTORS
potential
AFRICA
Adeyemi Akinlawon, Adeb Konsult
CANADIAN
DRILLING
Jeff Moss, ExxonMobil failure
Cam Matthews, C-FER Technologies
Well integrity
Khalid Zainalabedin, Saudi Aramco Hisham Saadawi, Ringstone Petroleum Consultants
WORLD CRUDE OIL PRODUCTION+‡ HENRY HUB GULF COAST NATURAL GAS SPOT PRICE‡
THOUSAND BOPD
6
O PEC JUL AUG SEP OCT
5 USD/million Btu
Algeria 1350 1350 1350 1350
Angola 1829 1833 1768 1618 4
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
DEC
2017
JAN
Kuwait1 2570 2570 2600 2650
Libya 310 250 310 550
Nigeria 1873 1913 1943 1988
Qatar 1537 1537 1477 1507
Saudi Arabia1 10670 10640 10600 10590 WORLD CRUDE OIL PRICES (USD/bbl)‡
UAE 3156 3186 3216 3196
Venezuela 2220 2210 2200 2190
2017
TOTAL2 35583 35695 35721 35985 JUN JUL AUG SEP OCT NOV DEC JAN
THOUSAND BOPD WTI 48.76 44.65 44.72 45.18 49.78 45.71 51.97 52.50
Africa 82 81 77 77 79 78 79
Asia Pacific 186 194 190 182 188 192 198
INDICES KEY
+
Figures do not include natural gas plant liquids.
1
Includes approximately one-half of Neutral Zone production. TOTAL 1481 1547 1584 1620 1678 1772 1918
2
Includes all current OPEC members.
3
The “Other” line item also includes Argentina, Australia, Azerbaijan, Brazil,
Colombia, Denmark, Equatorial Guinea, India, Kazakhstan, Malaysia,
Oman, Sudan, Syria, Vietnam, and Yemen. Monthly production from these WORLD OIL SUPPLY AND DEMAND‡
countries was listed individually in previous JPT issues. Ongoing work
on the US Energy Information Administration (EIA) website is disrupting
the regular updating of these countries’ production numbers. Additional MILLION BOPD 2016
annual and monthly international crude oil production statistics are
available at: Quarter 1st 2nd 3rd 4th
http://www.eia.gov/beta/international/.
† Source: Baker Hughes. SUPPLY 96.82 96.49 97.04 98.50
‡ Source: EIA.
Numbers revised by EIA are given in italics. DEMAND 95.36 96.09 97.46 96.94
Supply includes crude oil, lease condensates, natural gas plant liquids, biofuels, other liquids,
and refinery processing gains.
BENEFIT VISIT US
AT ICOTA
#617
HOUSTON, TX
OF RETROFIT
The two run solution was then mobilized. The lower straddle module High expansion straddle
was set below the water producing interval in the 5 ½” liner. The upper Single & multi run straddle
straddle module, in conjunction with Interwell’s unique stinger solution, Retrievable straddle
was successfully located into the latch system of the lower straddle
module. The upper straddle module was then set above the perforated
interval and effectively isolated the water producing zone.
The Multi-Run Custom HEX APS was designed and delivered ahead of the
customer’s tight operational schedule. The solution successfully isolated
the water producing zone and allowed the client to resume production.
YOUR
GLOBAL
PARTNER
www.interwell.com
Talent Everywhere
Janeen Judah, 2017 SPE President
The theme for my columns this year is years of the late 1970s and early 1980s retired in droves over
Risk and Reward because risk assess- the past 2 years. I see it when I talk to companies, when I walk
ment and decision making have fasci- around the halls at Chevron, and when I walk around down-
nated me for my entire career. The most town Houston at lunch—there are far fewer gray heads in
significant risks and rewards in our busi- our business. I’ll make an exception for management; many
ness are created by the judgments of company management are from that era, but in the techni-
the people who work here. People and cal workforce—mostly retired. So far, we have seen no conse-
their talent create upstream value. How quences because we are in a time of such low work activity. But
then do we generate the talent pool to create value in the when the upturn occurs—and it will happen—are we ready to
next upturn? move forward with a next-generation workforce?
If you ever doubt the impact that one person can make, let Management may think that the aged 55+ workforce will
me give two recent examples. come back in an upturn. I disagree. They’ve been through five
The biggest disaster in our industry—ever—was the Ma- downturns and worked especially hard over the past 10 years
condo blowout in April 2010. Like the day of the 11 September of frenzied activity. They are enjoying their time off, and near-
2001 attacks in the US, I’ll bet that most of you can remem- ly everyone I have talked to has no desire to return to full-time
ber when you first heard about the fire and crowded around employment. A minority is interested in part-time or project
a computer or television in horrified silence. I remember ask- work, but only on their own terms.
ing: Why didn’t someone use Stop Work Authority? Why didn’t A quick note to all of our valued SPE members who are en-
someone say something? joying their retirement: Please don’t retire from being active
In the recent movie depiction, Deepwater Horizon, actor in SPE! We have thousands of Young Professionals and stu-
John Malkovich played the evil manager who pushed forward dents who need your expertise and counsel. Your contribu-
despite signs of disaster. It was a cartoonish portrayal, but it tions are vital to the health and longevity of SPE.
vividly illustrated the impact that a single individual can have So, what will industry leaders do for this next upturn? We
on our operations. The result was a disaster on every level, al- need to be ready for the next-generation workforce. The
most destroying one of the largest companies on Earth. oil and gas industry is everywhere, so we need to find our
In contrast, on a positive note, there is the story of Steve talent everywhere.
Keenan, the experienced explorationist who led the team that Our business is much more global than it was when I
discovered Alpine High in the Permian Basin, the largest sin- joined. It’s no longer dominated by the “Seven Sisters” and
gle discovery in the US in many years. Apache’s management their mostly OECD-based workforce. National oil companies
gave Keenan and his team the freedom to think differently, have emerged in the past 40 years, and the top tier are ab-
and the results were spectacular. It’s ultimately people who solutely technically competitive with the old guard. The next
create (or destroy) value in our industry. tier isn’t far behind. The oil business is also more regional
Talent will never be more important than in the next up- than it was when I joined. Smaller, regional, or niche play-
turn. Companies have retired or laid off the most experienced ers have emerged, especially in the last upturn. These players
technical workforce in all parts of our industry: exploration are often technology-driven, but are almost universally more
and production, oilfield services, and engineering, procure- nimble than the old guard. They focus on a regional or tech-
ment, and construction. Just a few years ago, projects all over nology trend, like the US shale, Canada oil sands, or small-
the world were delayed because of a lack of people, not capital er African basins, and they are successful doing a few things
constraints. It won’t be long before that will happen again, and very well.
we will be looking for talent everywhere. These two trends—the Big Crew Change and the globaliza-
The “Big Crew Change” has happened. We’ve been talk- tion and regionalization of companies—create opportunities
ing about it for years and, in my view, it is mostly over. Peo- for talent everywhere. First, the next-generation technical
ple who, like me, came into the oil industry during the boom workforce will run the show in this next upturn. They think
The 2-year downturn in oil prices has been a challenge for oper- Alex Crabtree, Hess Corporation
ators and service companies alike, but operators appear to have Gunnar DeBruijn, Schlumberger
turned a corner, going by the most recent fourth-quarter earn- Mark Egan, Retired
ings. For the larger service providers, it may take longer to gain Mark Elkins, Retired
solid financial footing, but the chief executives of these firms are
Alexandre Emerick,
sounding more optimistic. Petrobras Research Center
After months of cost cutting and reassessment of projects, Niall Fleming, Statoil
earnings for the larger majors were positive. More stable oil pric-
Ted Frankiewicz, SPEC Services
es and OPEC’s recent production agreement point to a brighter year in 2017. Total boast-
ed a USD 548 million profit for the quarter, compared with a USD 1.6 billion loss in the Stephen Goodyear, Shell
fourth quarter of 2016. The company announced that it was ready to embark on new Omer M. Gurpinar, Schlumberger
projects, possible acquisition, and increased production. BP eked out a USD 72 million A.G. Guzman-Garcia, Retired
profit compared with a USD 2.2 billion loss in the year-ago period. Shell also reported Greg Horton, Retired
profits, although net revenue was down
John Hudson, Shell
from the previous year’s quarter. Shell
Morten Iversen, Karachaganak Petroleum
New JPT Website said it had “turned a corner” after paying
down debt and absorbing BG. Chevron Leonard Kalfayan, Hess Corporation
and Newsletter
posted its second straight quarterly prof- Thomas Knode, Contek Solutions
JPT has launched a new website and it and sees production growth this year Sunil Kokal, Saudi Aramco
will debut an electronic newsletter this amid cautious spending and cost con- Marc Kuck, Eni US Operating
month to keep members up to date trol. ExxonMobil, meanwhile, recorded
Jesse C. Lee, Schlumberger
with the latest upstream technology its lowest earnings in 2 decades and took
developments, trends, and news. The Douglas Lehr, Baker Hughes
a huge writedown on the value of some of
website (www.spe.org/jpt) features its upstream assets. Silviu Livescu, Baker Hughes
frequent updates and reports from Smaller operators, particularly those Shouxiang (Mark) Ma, Saudi Aramco
conferences, articles about technology involved in shale plays in west Texas, John Macpherson, Baker Hughes
applications, and interviews, including New Mexico, and other promising areas, Stéphane Menand, DrillScan
information not found in JPT’s print
plan more aggressive upstream spending
edition. The new website is more user- Graham Mensa-Wilmot, Chevron
this year. While many larger internation-
friendly than the previous one and also Badrul H. Mohamed Jan, University of Malaya
al plays still seem risky, activity in places
has a mobile-friendly design. Zillur Rahim, Saudi Aramco
such as the Permian Basin is soaring.
The monthly print edition will continue
to be mailed to all SPE members and will
Service companies are also seeing a Eric Ringle, FMC Technologies
continue to contain all of the features, better year compared with the previous Martin Rylance, BP plc
columns, and departments that members two, particularly for those involved in Robello Samuel, Halliburton
have come to read and rely on. North American operations, but still face
Otto L. Santos, Retired
The biweekly electronic newsletter some challenges. Many service providers
Luigi A. Saputelli, Frontender Corporation
will initially be sent to all members, but have begun renegotiating prices with cli-
ents, after slashing prices the past 2 years Sally A. Thomas, Retired
members must sign up for it on the JPT
website to continue receiving it. To ensure because of the steep fall in oil prices. “The Win Thornton, BP plc
that you continue to get the e-newsletter, direction [service companies] all need Xiuli Wang, Baker Hughes
please go to www.spe.org/jpt and enter to go is that we need to recover some of Mike Weatherl, Well Integrity
your name and email address in the Stay the pricing concessions that we’ve given,”
Rodney Wetzel, Chevron
Connected box on the right-hand side of Schlumberger Chief Executive Officer
the page. Scott Wilson, Ryder Scott Company
Paal Kibsgaard told the Wall Street Jour-
nal during an earnings presentation. JPT Jonathan Wylde, Clariant Oil Services
Lean Thinking is a generic version of exercise regularly, get plenty of sleep,” There have been measurable benefits in
the Toyota Production System, which lean is a simple concept difficult to live improved drilling times and reducing
was cobbled together just after World daily. Companies typically struggle the cost of bringing a well into produc-
War II. The system was how a resource- through distinct levels of understand- tion. But what are the benefits beyond
poor Toyota, staggering with broken ing and application of lean. Beginners drilling and completions, and applying
processes and a low-skilled workforce, dabble with lean tools in superficial lean across the extended value stream?
scrabbled through the harsh post-war ways that yield slim results. More devel- Put another way, think about your
economy to become one of the most oped users imbed lean more broadly corner McDonald’s: It fixes some glar-
successful and highly respected enter- across value streams and deeper into ing problems at their moneymaker, the
prises of the 20th century. The 1996 business systems to get greater returns. drive-through window. But how com-
book, Lean Thinking, introduced lean to Advanced lean thinkers go much fur- petitive is it if the orders are incorrect,
mainstream business. Since then, lean ther and get dramatic results by build- the assembly is sloppy, and it cannot get
has propelled operational excellence in ing lean into their culture. Like safety, ingredients to the right place at the right
construction, software development, lean for them is not another project to time? Oil and gas is still in the lean tools
healthcare, financial services, state gov- be implemented; lean thinking is part of era: Lean tools used in the business’
ernment, and more. how they do everything. most manufacturing-like processes. A
Lean Thinking is a philosophy that Oil and gas companies have begun windfall of returns await those who can
engages everyone in systematically solv- to use lean tools in isolated processes expand their scope of application and
ing problems. However, like “eat right, aiming to increase business efficiency. depth of understanding.
halliburton.com/geotech
Natural-Gas Dehydrator
ExxonMobil introduced its cMIST tech- ;ম
m]b|ub]_||_;Cuv|ঞl;-m7;;uঞl;u;tbu;v]oo7bm=oul-ঞomķ
nology, which dehydrates natural gas ]oo7]b7-m1;-m7]oo7|;-louh
using a patented absorption system
inside pipes and replaces the need for
!b7;u_;Ѵrv1u;v-1_b;;|_;0;v|u;vѴ|v0vrrouঞm]|;-lvķ
conventional dehydration-tower tech-
nology. This inline technology could be 7;;Ѵorbm]bm7bb7-Ѵv-m7_;Ѵrbm]1olr-mb;v|ol;;|or;u-ঞom-Ѵ|-u];|v
deployed in both land-based and offshore
natural-gas-production operations. The !b7;uvrrou|vou;mঞu;or;u-ঞomѴ;-v;1om|-1|v|oCm7o|lou;
technology, developed and field-tested
by ExxonMobil, more efficiently removes
water vapor present during the produc-
www.ridercorp.com
tion of natural gas, a process that reduces
corrosion and equipment interference.
cMIST reduces the size, weight, and cost Never miss a critical step
of dehydration, resulting in reductions of Optimize efficiency and control over the process
surface footprint by 70% and the over-
all dehydration system’s weight by half.
Asphaltenes represent the heaviest frac- tion and transportation of crude oils ly related to its hetero-atom content,
tion of crude oil, which are known to (Khvostichenko and Andersen 2009). with a higher hetero-atom content giving
precipitate when the crude is added to These tendencies result in reduced flow increased levels of polarity and a higher
aliphatic solvents such as n-pentane or or complete blockage of producing wells rate of aggregation (Hosseini et al. 2016).
n-heptane and yet remain soluble in light and surface equipment, including pumps,
aromatic solvents such as benzene or tol- pipelines, and separators. Experimental Device
uene (Gawrys et al. 2006; Borton et al. Currently, the only methods of treat- Our ultimate goal is to build a device
2010). They are characterized by highly ment are through the use of chemical dis- (Fig. 1a) that would remove asphaltenes
complex structures that contain multiple persants and inhibitors, which increase from crude oil near the point of pro-
aromatic rings and have a large hetero- the stability of asphaltenes to prevent duction, using electrokinetics. Thus, a
atom content (e.g., nitrogen, oxygen, and deposition. Once asphaltene deposition scaled-down device (Fig. 1b) was fab-
sulfur) and metal content (e.g., vanadium has occurred, running a “pig” through ricated and tested using a model oil to
and nickel) (Yarranton 2000; Hashmi the pipeline is often the method used to prove the concept and study some of
and Firoozabadi 2012). scrape the solids that accumulated on the the parameters that would influence the
Asphaltenes tend to self-associate on a walls of the pipe. It is known that asphal- design of a larger-scale device.
molecular level, depending on the com- tene molecules can be polarized, gain- An annular chamber was conceived
position, temperature, and pressure of ing an electric charge by introducing an where model oil flowed through the
the system. Precipitation of the particles electrostatic field (Hashmi and Firooz- annular space while applying a high volt-
out of solution results in flocculation, abadi 2012; Khvostichenko and Ander- age to create an electrostatic field so that
where they begin to deposit on hydro- sen 2009; Khvostichenko and Andersen the asphaltene electrodeposition under
phobic surfaces such as metal pipes and 2010; Hosseini et al. 2016). The polar- dynamic conditions could be investigat-
surface equipment used for the produc- ity of the asphaltene particles is direct- ed. A copper annulus was equipped as an
Motor
(a) (b)
Mixer
Voltage Unit
Valve 1
Splitter
Valve 3 Valve
Nitrogen Tank
Inner Copper Pipe Valve 2
Beaker
Teflon Pipe
Fig. 1—(a) An artist’s rendering of a prototype device that would electrokinetically remove asphaltenes from crude
oil near the point of production. (b) A schematic of the experimental setup that was used to prove the concept. The
charged annulus was composed of two copper pipes and was gravity-fed the heptol-70 from an elevated tank that was
stirred to ensure uniform asphaltene distribution in the oil. A nitrogen blanket was used to minimize volatilization of the
heptol-70 for safety. Source: Petroleum Engineering Department, University of Houston.
www.tomax.no
(a) (b)
(c) (d)
Fig. 2—Deposition onto the anode (inner copper pipe) after a dynamic, electrokinetic flow test for flow rates of:
(a) 1.5 ml/s (HRT=6.7 min); (b) 4.0 ml/s (HRT=2.5 min), and; (c) 5.5 ml/s (HRT=1.8 min). (d) The deposition layer
thickness is shown while being scraped from the anode after the 5.5 ml/s flow test to enable the mass to be measured.
Source: Petroleum Engineering Department, University of Houston.
electrode pair with a high-voltage power ing the solid asphaltenes to the hep- three figures that “the highest level of
supply, and the model oil was gravity-fed tol, whereas the precipitation approach deposition” on the negative electrode
into the charged annulus under varying involved dissolving the solid asphaltenes corresponds with the highest flow rate,
flow rates. The model oil consisted of a in toluene and then adding heptane. or “the shortest hydraulic retention
30:70 mixture of heptane and toluene For the asphaltenes and heptol-70 time (HRT).” Fig. 2d shows the deposi-
by volume, called “heptol-70,” to which used in our studies, our preliminary tion layer being scraped from the anode,
asphaltenes where added (2 wt%). tests showed no measurable difference and the gravimetric data confirmed the
Saturates, aromatics, resins, and in the polarity of the particles when pre- qualitative results.
asphaltenes (SARA) are all components of pared using the two methods, and we The dynamic tests were repeated
crude oil that can be precipitated from it. used the precipitation method for all the and confirmed an inverse relationship
Our earlier work applied electrokinetics dynamic tests. between HRT and amount deposited
under “static conditions” to the asphal- The small-scale device consisted onto the negative electrode—i.e., higher
tene fraction only, and concluded that the of a copper annulus and was charged flow rates resulted in a higher degree of
asphaltene particles could be attracted to to create an electric field; the inner deposition. These results imply that the
the anode (Hosseini et al. 2016). pipe being charged negatively and the deposition of the asphaltenes is depen-
Work by Khvostichenko and Anderson outer positively. dent upon the amount of particles that
(2010) indicated that resins may neu- come in contact with the charged elec-
tralize the asphaltene charge, affecting Testing trode, and also indicates that the effi-
the electrodeposition of asphaltenes. With the heptol-70 containing a 2-wt% ciency of the electrokinetic process is
To understand whether the other SARA concentration of asphaltenes in the ele- high for the range of flow rates tested. We
components found in crude oil would vated reservoir and the desired flow rate did observe some settlement of asphal-
interfere with the electrokinetic process achieved, the power supply was turned tene particles at the bottom of the cham-
under “dynamic conditions,” saturates, on for 5 minutes. The average flow rate ber that we attribute to the design of the
aromatics, and possibly resins were used was measured while applying voltage. device, where a small lip surrounding the
along with the asphaltene particles in the Once completed, the anode was removed exit acts to trap some of the particles.
heptol-70. While it is unlikely that the and photographed, and the asphaltenes
precipitates used in our study contained were scraped from the anode and col- Data Gathered
resins, because of the procedures that lected so that the mass removed could The data gathered at this early stage of
were used, we cannot rule out the possi- be weighed. research suggest that the particles, a mix-
bility at this early stage of the work. Three flow rates were used while keep- ture of asphaltenes, saturates, aromat-
In addition, we conducted preliminary ing the duration of the experimental run ics, and possibly resins, can be deposited
tests to compare two procedures for pre- and the voltage constant at 5 minutes and onto an electrode and removed from the
paring the model oil, referred to as a 5,000 V, respectively. Figs. 2a through flow stream as a means of addressing a
“dissolution” or “precipitation” meth- 2c show the deposition of asphaltene critical flow-assurance problem.
od (Khvostichenko and Andersen 2010). particles onto the inner copper pipe for Future experiments are planned to
The dissolution approach involved add- the three flow rates. It is clear from the explore the limitations of the process in
ADVANCED
TECHNOLOGY
VALVE
E&P NOTES
As big data analytics becomes more inter- the drilling crew which of the next cut- stuck pipe. “From that, we can calcu-
twined with the operations of some of ters is most likely to fail next. Presum- late how much money [the operator] lost
the biggest names in upstream, this tech ably, such information could influence a because of certain actions and symp-
space is beginning to deliver the industry driller to change tactics to extend the life toms,” said Priyadarshy.
some compelling applications. The most of the bit, and therefore reduce the num- The next phase of this project is to
recent data were highlighted in a discus- ber of time-consuming trips out of a well build a predictive model that begins look-
sion led by Satyam Priyadarshy, Hallibur- to retool. ing for symptoms and actions before the
ton’s chief data scientist, who declared The company is also building an “earth drilling report is even written up. Priya-
that big data analytics “will transform model” that will simulate the earth’s sub- darshy said this will involve creating a
the industry.” surface using industry data, and has 150 way for the drillers to input more mean-
That transforma- PhDs from the fields of geochemistry, ingful observations than are currently
tion is to be real- fossil research, and geology working on included in a typical report.
ized by using ana- this initiative.
lytics to identify Get Educated and Find Help
and eliminate hid- Fixing Stuck Pipe Priyadarshy also remarked on the factors
den inefficiencies. In regard to solving big problems, Halli- he sees holding back the industry from a
According to Priya- burton may be close to solving the drill- faster uptake of analytics. Of his biggest
Priyadarshy darshy, that means ing issue known as stuck pipe. Among concerns is that too many people in the
that solving known the oil field’s oldest sources of cost over- industry simply think of big data as “a lot
problems is not enough; to achieve a runs, stuck pipe involves a situation in of data,” which he called a “fundamentally
higher order of innovation, oil and gas which a drillstring cannot be pulled flawed definition.”
companies must discover and answer from the borehole without leading His definition: “Big data is about all
problems they never knew about in the to damage. the data of the business, which means
first place. Using a historical data set as an exam- the historical, current, and future data
Speaking at the first meeting of the ple, Priyadarshy said analytics could that you will generate.” This understand-
SPE Gulf Coast Section’s Data Analyt- have saved an anonymized operator ing falls in line with a concept that might
ics Study Group in Houston, Priyadarshy USD 17 million out of the USD 22 mil- be described as total-analytics, in which
explained how Halliburton analyzed data lion it spent on stuck pipe issues during a companies allow the entirety of the their
from 5,000 wells and found that a spe- multiwell drilling campaign. “That is the data to be placed under the microscope
cific pump model was suffering from an value of the data that was just sitting in in order to drive out wasteful workflows
inordinate amount of downtime. Hidden somebody’s warehouse,” he said, empha- and unnoticed problems
within the data was the simple fact that sizing a point that though the industry “The reason [industry workflows]
the pump was not well-suited for one par- collects big data, it rarely uses it. are inefficient is because whatever
ticular climate zone. Priyadarshy said the program took inefficiencies we knew about, we have
“Now we can actually predict that a only 6 weeks to train and works by taken out, and beyond that, we can’t see
certain pump, of a certain manufactur- analyzing what he called “drilling lan- because we haven’t yet looked for them,”
er, should never be sent out to a certain guage.” This variation of natural language he explained, before adding, “The only
area,” said Priyadarshy, noting that the machine learning takes the unstructured way you can actually be innovative is by
cost of the analytics platform has already text data found in the human-written looking holistically at the problems of the
been recovered by the revenue earned drilling reports and connects the dots in industry through big data analytics and
from keeping the pumps running longer. a way that is seductively simple. by building new intelligence that is based
Priyadarshy described another ana- In this case, the dots are represent- on the hidden inefficiencies.”
lytics program that, when told a specific ed as three categories on a bar chart: Aside from the steep learning curve,
cutter on a drill bit has failed, will show symptoms, actions, and the event—a the oil and gas business is facing a major
Could it be that the upstream industry tory, having seen installation prices drop Renewables have also been discussed
has an unlikely ally in renewable energy? by 66% since 2010, according to the as a way to lower offshore facility costs.
The two sectors are often framed as being American Wind Energy Association. The Norwegian certification society DNV
at odds—one is supposed to eventual- Several industry majors have attempt- GL presented a paper (OTC 25284) at
ly replace the other, after all. But in the ed to fit renewable energy into their last year’s Offshore Technology Confer-
meantime, the falling price of renewable portfolios, but as Engel-Cox pointed out, ence in Houston that showed wind tur-
technologies has opened the door for oil none have made it past the pilot stage. bines could be used to power injection
and gas companies to reduce the cost of “A lot companies have been in the solar operations of 45,000 B/D of water into
electrifying their fields, according to one business multiple times, a number are one of Norway’s remote offshore reser-
US energy expert. looking at biofuels, and few at wind,” she voirs. There are other ideas being tossed
“In many locations, solar and wind are explained. “But in the past, these efforts around that include using wind or solar
now cost-competitive with other forms have been more about developing a new power for natural gas pipeline compres-
of electricity generation, so it just makes product-side of the business—not neces- sion and for water desalination.
economic sense to be incorporating them sarily about integrating renewables into Regardless of which type of renew-
into your industrial operations.” said Jill their operations.” able-integration project upstream opera-
Engel-Cox, the deputy director of the US Her conclusion is that more projects will tors decide upon, Engel-Cox said the first
Joint Institute for Strategic Energy Analy- achieve commercial success if a simpler step toward success will depend on the
sis (JISEA). approach is taken, such as one that focus- quality of economic models and site-spe-
Founded by five universities and the es on using renewables simply as a power cific studies that tell them things such as
US Department of Energy’s Nation- source for energy-intensive machinery how long the sun shines or how hard the
al Renewable Energy Laboratory, JISEA such as pumps and compressors. wind blows.
is tasked with researching transdisci- Among the few examples of field-inte- “I don’t think we know what the most
plinary global energy issues, including grated renewable projects is Chevron’s profitable opportunities are yet,” she
how to exploit synergies between renew- enhanced oil recovery (EOR) project in emphasized. “But I think looking at a
ables and hydrocarbons. California that used concentrated solar variety of options is the best idea. They
The bad news for the champions of this power to generate steam for reservoir will have to be looked at in the context of
concept is that there is little momentum injection. Though the solar-to-steam each facility, the type of power demand it
within the industry to implement such project was seen as groundbreaking upon has, and where it is physically located.”
integration. The good news is that many startup, construction costs far exceeded
of the early failures were attributed to the the budget and it was shut down in 2014 For Further Reading
cost of renewables; a factor that is quick- after 4 years of operation. SPE 169745 Construction, Operations
ly fading away. This same concept is being revived and Performance of the First Enclosed
Bloomberg data published in Janu- half a world away in sunny Oman. Petro- Trough Solar Steam Generation Pilot
ary showed that the cost of solar power leum Development Oman, a government for EOR Applications by Daniel Palmer,
has dropped by 62% since 2009, mak- majority-held production company, is and John O’Donnell, GlassPoint Solar.
ing new installations more affordable working with GlassPoint to build a solar http://dx.doi.org/10.2118/169745-MS.
than coal power in several countries. The facility that will generate 6,000 tons of OTC 25284 Wind-Powered Subsea Water
report also projected that solar power steam per day for EOR injection. The proj- Injection Pumping: Technical and
could become cheaper than coal on a ect calls for the construction of four glass- Economic Feasibility by Johan Slätte
global basis by 2025. In the US, wind house-solar facilities, the first of which is and John Sandberg, DNV GL, et al.
power generation is on a similar trajec- to begin operating sometime this year. http://dx.doi.org/10.4043/25284-MS.
Fossil fuels will probably be at the fore- oping countries; encourage US fossil fuel maker” and said that will be “really, real-
front of energy and environmental policy exports; and shift environmental policy ly impactful from an energy standpoint.”
under the new Trump Administration, away from a central emphasis on carbon Trump has talked extensively about
said Charles D. McConnell, executive dioxide (CO2) emissions. renegotiating the North American Free
director of Rice University’s Energy and “The time I served in Washington, envi- Trade Agreement (NAFTA). “When we
Environment Initiative, in a recent talk ronmental responsibility became equiva- first did NAFTA, energy with Mexico was
hosted by the Norwegian Consulate Gen- lent to a fixation, a singular focus, on not on the table—politically not on the
eral in Houston. carbon dioxide,” McConnell said. “And I table, not allowed to be,” McConnell said.
McConnell served 2 years as an assistant have got to tell you, there is a lot more to “Could it be on the table? Sure, a huge
Secretary of Energy in the Obama Admin- environmental responsibility than CO2 impact, not just for the US but also for
istration before taking his current posi- emissions. People need food, water, and Mexico. So that is an example, I think,
tion. He has spent 35 years in the energy access to energy.” of looking at the trade agreements and
industry, which includes a 31-year career Approximately 1.4 billion people are looking at them in a different way.”
at Praxair followed by serving as a vice without energy in the form of electrical Technology investment is another area
president at Battelle Energy Technology. power, and this is in areas with the most where the Trump Administration is likely
Notwithstanding new US President population growth and that are expect- to shift direction.
Donald Trump’s famous unpredictabil- ed to produce 90% of global economic “If you look around the world in the
ity, McConnell suggested that the admin- expansion in coming decades, McConnell last 15 years, we have probably spent
istration would emphasize oil and gas said. For these populations, some of the over USD 20 trillion developing renew-
development; stress the continuing need most critical needs are for energy sources able technology,” McConnell said, refer-
for oil, gas, and coal in the world’s devel- that can ensure clean water and eliminate ring to investment from global sources. “I
indoor air pollution from cooking over am not suggesting that is a bad thing, but
poorly vented fires. Increasing the avail- it is a stunning number, isn’t it?
ability of fossil fuels to generate electrici- “During that same period of time, we
Changing Your ty will bring a major improvement to peo- have invested USD 20 billion in fossil
ples’ lives in these areas, he said. fuel technology,” he continued. “And yet
Address? 80% of our energy globally is fossil fuels.
Let SPE know.
Energy Links to Water, Food And guess what the IEA [International
“The connection of that energy/water Energy Agency] projects by 2050? We
+1.972.952.9393 or
nexus, the energy/food nexus is incred- will double the world’s demand for ener-
+1.800.456.6863 (toll-free) ibly powerful,” McConnell said. “I believe gy, and it will still be 80% supplied by
it will be at the top of the list of this fossil fuels. And yet we are spending all
new administration.” our money on renewables. I don’t think
Elsewhere in the world, there are uncer- that is going to continue to happen in
Update Your tainties over transnational energy sup- this administration.”
plies that could be curtailed or used as a McConnell said he did not believe that
Member Profile policy lever for political reasons, he said. the US would drop out of the COP 21
“It is all about energy security,” global agreement on reducing climate
http://www.spe.org/
McConnell said. “The rest of the world change, as some people have speculated.
members/update
needs to know that the United States is “Is Trump going to roll back every-
going to continue to develop its resourc- thing environmentally? No, he won’t,”
es. We have no reason not to be the lead- McConnell said. “That is not going to
er in the world for energy, and we are happen, primarily because most of the
SPE Benefits becoming that.” American companies—major American
Commercial arrangements, how ener- companies—have a global footprint. You
Discover the possibilities. gy is traded, will probably be a focal can’t have requirements in the US that do
point of Trump Administration policy. not meet global expectations. Rex Tiller-
http://www.spe.org/
“Trump is likely to play to what he sees son [the new US Secretary of State and
members/benefits
as the advantage of the US,” McConnell former ExxonMobil chief executive] will
said. He characterized Trump as “a deal- tell him that in spades.” JPT
IS NOW COMPATIBLE UP TO
350K PPM
TOTAL DISSOLVED SOLIDS
Apply now.
www.aramco.jobs/jpt
Bill Grieser, and Travis Bachman,
MESH REPRESENTS THE LARGEST-SIZE PARTICLE WITHIN A SAND GRADE, Halliburton. http://dx.doi.
WHICH INCLUDES SMALL ONES
org/10.2118/184863-MS.
SPE 174060 Application of Micro-
Mesh Microns Inches Millimeters
Proppant To Enhance Well Production
30 595 0.03 0.841
in Unconventional Reservoirs:
40 400 0.0165 0.4
Laboratory and Field Results
50 297 0.012 0.297
by Jeff Dahl, Devon Energy;
60 250 0.0098 0.25
Philip Nguyen, Ron Dusterhoft
70 210 0.0083 0.21
et al., Halliburton. http://dx.doi.
100 149 0.0059 0.149
org/10.2118/174060-MS.
270 53 0.0021 0.053
SPE 184858 An Experimental
325 44 0.0017 0.044
Investigation of the Conductivity
400 37 0.0015 0.037
of Unpropped Fractures in Shales
530 25 0.0009 0.025
by Weiwei Wu, Mukul Sharma et al.,
The University of Texas at Austin.
Source: Flocon Filters. http://dx.doi.org/10.2118/184858-MS.
SPE 174946 The Role of Induced
Un-propped (IU) Fractures in
ent,” Parker said. Rapid drilling and com- For Further Reading Unconventional Oil and Gas
pletions are based on complex routines SPE 184863 Enhancement of Well Wells by Mukul Sharma and R.
that companies are reluctant to change, Production in the SCOOP Woodford Manchanda, The University of
but he added, “Some operators are now Shale Through the Application of Texas at Austin. http://dx.doi.
looking at it seriously.” JPT Microproppant by James Calvin, org/10.2118/174946-MS.
Motivate
Inspire
Educate
Nominate
Do you have colleagues who are authorities in their fields and experienced public speakers?
If you do, consider nominating one or more of them for the Society of Petroleum Engineers
Distinguished Lecturer Program.
Learn more about the program at www.spe.org/go/NomDL.
Nominations are accepted until 15 March.
The SPE Distinguished Lecturer Program is funded by the SPE Foundation, Offshore Europe,
AIME, and companies that allow their professionals to serve as lecturers.
he mostly widely reported production In comparison, the output of conden- being a good correlation, from 0.6 to
T data proved to be a bad indicator of
future output.
sate wells over 30 days is a useful indica-
tor, though 90 days offers a better fit with
0.85 a moderate correlation, and below
that little correlation.
A study presented at the confer- the longer-term data. A short-term time frame is signifi-
ence considered whether the widely The surprises in the study’s findings cant because the presentation showed
reported 30-day production total is a include production data that were both that output over 30 days is the mea-
good way to evaluate a well compared better and worse than expected, with a sure normally reported to investors and
with less frequently offered longer- significant number on the upside, he said. the media. The impact those reports can
term numbers. The paper used data from more than have on a company’s stock price puts a
“Production at 30 days tends to be an 600 gas and condensate producing wells premium on the importance of early well
unreliable estimate,” said Edward Ifejika, in the Utica formation over a 2-year peri- performance, which can influence how
a field operations engineer for Total who od. It used a linear regression analysis the wells are completed and operated in
did the research while he was a student to test whether there was a correlation ways that may alter later results. The first
at Texas A&M University (SPE 184817). between actual production at earlier and month of production is also a time when
The number after 90 days offers a much later time periods—30, 60, 90, 180, 360, output may still be influenced by well
better estimate of future gas produc- and 720 days. cleanup after fracturing.
tion and the improvement from there The result of the calculation is called While the 30-day number is a pret-
is limited. R2, with a number greater than 0.85 ty good fit for the 90-day results (R2 of
Summary of the analyzed time intervals for wellhead gas and condensate. Source: Paper SPE 184817.
The image on the left from a well in the Bone Springs formation in west Texas was the first one created using
electromagnetic imaging, and the image on the right shows reprocessed data from that 8,000-ft deep lateral revealing
far more detail. Source: Paper SPE 184880.
said Terry Palisch, global engineering magnetic-coated particles. Getting good “The model suggests there is a
director for Carbo, adding that, “the rest data requires identifying noise, which threshold of proppant concentration”
is reasonably off-the-shelf technology.” can range from nearby power lines to imaged, Palisch said, adding, “The
Carbo has spent years developing the cows grazing in the area, and finding propped area is larger.”
sophisticated mathematical formulas ways to avoid, or remove it. The company recently complet-
required to interpret faint changes in Based on what it learned from its first ed the field work for a third test of
electrical and magnetic signals. In the test, Carbo increased the percentage of the system and is now processing that
year since Carbo’s paper revealed its receivers picking up electric and magnet- data, he said. It is working to increase
work, the company has improved the ic signals. While the two sorts of electro- the power of the electric current used
clarity with more data. It has learned to magnetic waves offered similar pictures, to stimulate the proppant and use
better locate the receivers on the sur- adding receivers to pick up magnetic data more of the signal and find better
face to detect the faint electric and mag- may offer more volume and detail, per- ways to reduce noise to create more
netic signatures created when it runs haps showing areas with lower concen- detailed pictures showing lower prop-
a current from the well through the trations of proppant. pant concentrations.
igher life expectancy, plentiful food, “The shale energy industry could the- “The moral goal that we accept with
H and soaring gross domestic prod-
uct are among the benefits that much
oretically have made a very exciting case
about how ‘We are going to get all this
green energy is based on the standard of
minimal human impact,” he said. “I want
of the world’s population has enjoyed energy out of previously useless rocks,’” to ask, ‘Is that a good standard; should
since widespread use of fossil fuels began he said. “But the industry did basically we be judging our actions by how little we
more than a century ago. They are also none of that.” impact nature?’”
the central pillars to Alex Epstein’s the- As a consequence, opponents of If the answer throughout humanity had
sis of why fossil fuel production, and hydraulic fracturing filled an outsized been a definitive yes, Epstein said, then
the engineering involved, equates to a share of the information vacuum, exem- modern cities and most people over the
moral obligation. plified by the controversial and popular age of 30 would not exist. These points
Epstein, founder of the for-profit film Gasland—a documentary that the are drawn from his philosophy of human
think tank Center for Industrial Progress industry has spent years trying to counter. flourishment. In contrast to natural con-
and author of The Moral Case for Fossil Epstein said that the oil and gas indus- servation, Epstein’s view on human flour-
Fuels, was the featured speaker of this try’s challenge of gaining greater public ishment holds that the chief moral cause
year’s conference. favor has been made even more difficult of mankind is to achieve progress by alter-
Addressing several hundred industry by the emergence of renewable energy ing the natural world to its own benefit.
professionals, Epstein used the rise of the technologies, which he views as imprac- For him, this is where common ground
North American shale sector to highlight tical apart from hydroelectric power may be found with industry detractors.
how, despite making possible the many generation. Among his problems with Epstein said that when people who dis-
advantages of living in the modern age, renewables is that they are generally cast agree on fossil fuel usage can agree on the
the wider oil and gas industry is losing as “good” while fossil fuels are framed by need to propel humans into the future,
the public relations battle. many as “evil.” “then we can have a discussion, because
Fcanoronbeallindifficult
of the upstream activity going
the Permian Basin of Texas, it
to assess which horizon-
ates will also help boost daily production
and long-term recoveries, placing more
revenue into the University Land endow-
is a University Lands program, Spath said
it has the “intimate” geologic and local
surface data usually only held by oper-
tal well strategies are working best and ment—which collected USD 850 million ators along with the regional but less
why. The primary reason is operators in in oil and gas royalties last year. detailed data generally held only by con-
Texas are not required to share as much At the helm of sultants and service companies.
detailed drilling and completion infor- this new orga- “We think of ourselves as that cross-
mation as they are in other major oil-pro- nization is Chief section where we have access to all uni-
ducing states such as North Dakota, home Executive Officer versity lands’ operators and the funda-
to one of the shale sector’s most revered Jeff Spath, the mental, detailed geologic knowledge of
public databases. 2014 SPE Presi- the subsurface,” which will be used to
But with the recent formation of the dent and former identify which drilling and hydraulic
Texas Oil and Gas Institute (TOGI), it Spath executive advisor fracturing trends are the ones to follow,
is expected that the Permian will soon at Schlumberger. he said.
become a much more well understood Speaking at the conference, Spath While not all the data TOGI has access
shale play. TOGI was established in 2015 explained that TOGI has a unique posi- to are allowed to be shared publicly, such
by the regents of the University of Texas tion when it comes to visibility of the as proprietary seismic data, Spath said
(UT) system, which manages oil and gas oil field. his long-time participation with SPE has
mineral rights across 2 million acres of State law requires that operators work- inspired him to ensure that the program
land in west Texas. ing on University Lands share their well “publishes everything that is publishable”
A nonprofit organization, TOGI information with the state. Because TOGI in industry technical papers.
describes its mission as one focused on
improving the value of the university
lands for the benefit of both the UT system
and the Texas A&M University system—a
co-beneficiary of the lands program.
Known as University Lands, these non-
contiguous areas were granted to the two
universities through a series of legislative
acts that began in 1839, 6 years before
Texas would go from an independent
republic to the 28th US state. Today, there
are more than 280 companies that oper-
ate 9,600 wells on this land, the largest
of them being Pioneer Resources, Apache
Corporation, and Devon Energy. Output
from the fields within University Lands is
about 42,000 BOE/D.
One objective for TOGI is to help devel-
op undergraduate and graduate students
Highlighted in orange, University Lands manages both surface and mineral
into skilled petroleum engineers and geo- interests across more than 2 million acres of land in 19 counties. Oil and gas
scientists. However, if all goes according royalties represent the overwhelming majority of revenue from these lands,
to plan, the learnings that TOGI gener- which were tallied at USD 850 million last year. Source: University Lands.
hen hydraulic fracturing arrived hold the parts together, forming pil- Aramco is working through a long-term
W in Saudi Arabia, there was some-
thing lacking.
lars that hold open channels for the
gas to flow.
program to determine the best meth-
ods for developing its unconventional
While the reserves of tight gas are These clumps are stronger than sand gas reserves.
bountiful and sand covers the des- alone, and when some of the grains break, But not all sand dunes will suffice.
ert, Saudi Aramco has had to import there is no evidence the fines escape. While some sand has met ISO and API
ceramic proppant to hold open the frac- The company compared the results requirements, some could not stand up
tures created. Sand dunes in its des- from 15 stages in a 4,500-ft lateral of to the 8,000-psi closure pressure. The
erts provide a seemingly endless sup- a well. When the three stage groups sand described in the technical paper
ply of sand, but when the national oil were compared, the stages using the came from three dunes that had eas-
company studied those grains of sand, local sand outperformed ceramic prop- ily accessible sand and high quartz
it found that they were “nice and pant stages using a similar design, (95%) content. JPT
round but not very strong,” said Kirk and were not far behind the last four
Bartko, a senior petroleum engineer- stages near the heel of the well that For Further Reading
ing consultant for Saudi Aramco, dur- were fractured conventionally using SPE 184817 Optimum Production
ing a presentation at the conference more proppant. Metrics To Predict Unconventional
(SPE 184823). When production from the test well Well’s Long-term Performance
It was a second-grade proppant, was compared with another well 5 km by E. Ifejika, B. De Cumont,
which was not quite up to the demand away fractured with ceramic proppant N. Kashani, Total E&P. http://dx.doi.
of wells typically 10,000 ft deep where only, the cumulative production after org/10.2118/184817-MS.
proppant faces closure pressures of 110 days was within 3%. “There has not SPE 184880 Far-Field Proppant
about 11,000 psi. The likely result would been a loss in rate from the two wells. Detection Using Electromagnetic
be sand grains crushed into fine parti- The pillar structure is sustaining what we Methods—Latest Field Results
cles likely to block the flow of gas. hoped,” Bartko said. by T. Palisch, W. Al-Tailji, L. Bartel,
The technical paper reported a work- For Saudi Aramco, this represents a Carbo Ceramics et al. http://dx.doi.
around for that problem that includ- huge potential savings over importing org/10.2118/184880-MS.
ed pumping a mixture of 70% dune high-performance ceramic proppant for SPE 184823 From the Backyard Dune
sand and 30% ceramic proppant the job. to Fracturing a Highly Tectonically
and fibers developed by Schlumberg- “We have the large dunes and the cost Complex Formation in Saudi Arabia
er. The sand was incorporated in the of our completions will go down tremen- by K.M. Bartko, I.H Arnaout,
service company’s channel-fractur- dously” if these test results lead to wide- K.S. Asiri, Saudi Aramco et al.
ing method, which uses the fibers to spread use of local sand, he said. Saudi http://dx.doi.org/10.2118/184823-MS.
nov.com/drillstring
© 2017 National Oilwell Varco | All rights reserved
MANAGEMENT
C en
targets
ific ss
ro ga
n
io
This is especially difficult in hierarchi-
ss g
pl e
at
m oc
-fu em
si Pr
cal organizations where decisions and
nc en
tio t
processes usually cascade from the top.
na
l
However, this can be overcome by push-
ing decisions to the appropriate levels KPI Continuous Supply chain
and by simplifying the decision-making monitoring Improvement management
process to be able to fast-track decisions
in response to business needs.
ps
Si
hi
x
ns
Si
tio
gm
Implementing the System
la
a-
Re
6σ
BHP Billiton’s unconventional basins Cost reduction NPT
were ideal candidates to implement a initiatives mitigation
Lean Well Manufacturing System. For
this reason, by having full leadership sup- Time and cost
port and adopting lean tools and pro- variability
cesses, it was possible to change the com-
pany mindset and become a continuous Fig. 2—The continuous improvement cycle. NPT: Nonproductive time; KPI: Key
improvement organization. Eight ele- performance indicator.
ments contributed to the transformation
(Fig. 2). ed in renegotiating contracts and rates ner mindset in which these companies
as market conditions changed. In addi- became integral players in the opera-
1. Lean training. tion, it engaged with vendors and tion and to identify new service pro-
suppliers to create an extended part- viders that could deliver services and
Educating and developing leaders, office,
and field personnel in lean methodology
was the starting point. Several dedicated
training sessions were performed with
participation of every key player involved
in the well delivery process and projects
assigned to put in practice the concepts The Petrolink Group of Companies provides a comprehensive data solution from
learned during these sessions. acquisition, aggregation and visualization to processing, delivery and analytics.
2. Target definition.
Hydraulic Fracturing
Zillur Rahim, SPE, Senior Petroleum Engineering Consultant, Saudi Aramco
Hydraulic fracturing is widely used all creating longer and planar fractures for fracture conductivity is provided by the
over the world, and in the Middle East improved production. open areas within the system. The crush-
in particular. Although superhigh-quality Diagnostics using production logging, ing of pillars by high in-situ stress and the
reservoirs do not need stimulation, the temperature logs, and injecting and mea- friable quality of sand and consequent
current expansion of reserves is mostly suring nonradioactive tracers have shown release of fines are controlled and con-
in newly found tight and unconventional expected production contribution and tained by the use of resin during the treat-
resources that require extensive multi- increase from target intervals, in both ment. This technology is also designed
stage fracturing for commercial produc- acid- and proppant-fracturing treatments. to reduce fracturing cost by eliminating
tion. Saudi Aramco embarked on stimula- The superiority of NDSs lies in their natu- more-expensive intermediate-strength
tion and fracturing of wells at the begin- ral formulation, simultaneous use in near- proppant (ISP) or high-strength prop-
ning of its nonassociated conventional well and far-field applications, robustness pant (HSP) during channel fracturing.
gas development endeavor and has been in creating barriers and diversions, and Sand cannot substitute for ISP or HSP in
expanding and improving continuous- easily degradable nature and flowback. conventional fracturing because of its low
ly in the areas of fracture design, use of Among other technology being devel- compressive strength and friable nature.
novel materials, and field implementa- oped and adopted in Saudi Arabia is the The economic advantage achieved by
tion. With newly discovered fields and use of seawater-based fracturing fluids that the use of new technologies will help
the extension of existing areas, the chal- will save the more-valuable fresh water. expansion, exploration, drilling, and frac-
lenges related to reservoir heterogene- The laboratory experiments and the few turing in new frontiers and territories.
ity, tight rock, layered systems, and field field applications to date have shown sta- The world’s increasing energy demand,
maturity have led to innovative ideas and bility at high temperatures and pressures, which, for the most part, is hydrocarbon
to testing and application of technologies. compatibility with formation fluids and dependent, requires growth and advance-
The use of biodegradable novel diver- fracturing additives, and resilience against ment in the oil and gas industry, increas-
sion systems (NDSs) has allowed fractur- scale formation. The use of seawater will ingly environmentally friendly practices,
ing multiple intervals in a single-stage serve the environment tremendously and the use of more natural resources, and
operation, thereby reducing time and sig- will provide substantial economic bene- the application of innovative ideas and
nificantly cutting operational cost. On fits in remote areas where access to fresh new technology that will reduce develop-
the basis of petrophysics and reservoir water is challenging and costly. ment and management cost and increase
and geomechanical characteristics, mul- Saudi Aramco also has initiated the efficiency and effectiveness. JPT
tiple mesh sizes of an NDS are careful- use of local sand as proppants combined
ly designed and used to divert fluids in with the channel fracturing technique,
the wellbore, ensure simultaneous treat- where proppants serve as pillars support- Recommended additional reading
ment of lower-quality intervals, and con- ing and maintaining the induced frac- at OnePetro: www.onepetro.org.
trol excessive leakoff inside the fracture, tures, keeping them open while the actual
SPE 180207 Effects of Hydraulic Fractures
on the Treatment of Condensate by Huff ’n’
Puff Gas Injection in Eagle Ford Shale
Zillur Rahim, SPE, is a senior petroleum engineering consultant
by S. Yang, University of Calgary, et al.
with Saudi Aramco’s Gas Reservoir Management Department.
With 30 years of industry experience, he heads the technical SPE 181353 Best Practices and Lessons
team and is responsible for hydraulic fracturing, well comple- Learned From More Than 1,000 Treatments:
Revival of Mature Fields by Hydraulic
tions, assessment and approval of new technologies, and train-
Fracturing in Khalda Ridge, Egypt’s
ing and development of local talent. An active member of SPE, Western Desert by Mohamed Salah,
Rahim has authored more than 100 technical papers and has Khalda Petroleum Company, et al.
participated as cochairperson, session chairperson, technical
SPE 184840 Innovative Diversion
committee member, discussion leader, forum coordinator, and workshop organizer
Technology Ensures Uniform Stimulation
for various Middle East and international SPE events. He holds a BS degree from Treatments and Enhances Gas Production:
L’Institut Algérien du Pétrole, Boumerdès, Algeria, and MS and PhD degrees from Examples From Carbonate and Sandstone
Texas A&M University, all in petroleum engineering. Rahim serves on the JPT Editorial Reservoirs by Zillur Rahim, Saudi Aramco,
Committee and can be reached at zillur.rahim@aramco.com. et al.
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
200 350
Pressure (psi)
Pressure (psi)
7,500 300 7,500
150 250
5,000 200 5,000
100
150
illustrated in Fig. 1b. The application finity for the released crosslinker. This to bind to the pressure-released cross-
of high pressure and temperature to breaking would allow the gel to main- linker. Those chemicals fall into three
borate-crosslinked polymers may alter tain the low viscosity and encourage im- categories that have known affinity for
the borate anion, inactivating the cross- provement in retained conductivity of boron: inorganic metal oxides (includ-
link and allowing the new form of boron the proppant pack. ing clays), activated solids, and polymer-
to undergo further reaction. The subse- ic solids.
quent proposed breaking reaction has Candidate-Material Screening. The The initial phase of the current qual-
the released crosslinker reacting irre- current work identified a number of ification gauged the feasibility of the
versibly with a chemical with a high af- chemicals with the established potential candidate breakers to react with the
Introduction
Rock mechanics research has shown
that hydraulic fractures propagate per-
pendicular to the minimum horizontal
stress in a normal fault environment, Fig. 1—A horizontal well with 10 transverse fractures and 500 ft of fracture
creating transverse fractures. This oc- half-length.
curs if the perforations are aligned with
the preferred fracture plane, which, in were used. None of the previous work un- horizontal wells vs. longitudinally frac-
this case, is the maximum horizontal dertook extensive integrated completion tured horizontal wells in tight sands and
stress. However, the debate has centered and reservoir simulations that modeled unconventional reservoirs with stress-
on whether transversely fractured hori- multiphase flow in transversely fractured dependent permeability were conducted
zontal wells or longitudinally fractured vs. longitudinally fractured horizontal using three different reservoir-simulation
horizontal wells are appropriate and best wells. This study incorporated the effect models. The first reservoir-simulation
practice in a given area and for a given of non-Darcy flow, adsorption gas, the model was for a dry-gas reservoir (in
reservoir permeability. relative permeability effect on fluid flow contact with water) and produced only
The motivation for conducting this re- in the fracture, and the effect of stress- gas and water. The second reservoir-
search came out of the realization that dependent permeability on fracture con- simulation model was built for a black-oil-
all previous studies that looked into the ductivity, which were missing in previ- type reservoir (undersaturated), which
performance comparison of transverse- ous studies. honored accurate reservoir-fluid prop-
ly vs. longitudinally fractured horizontal erties for Permian Basin oil. The third
wells were limited in scope either by the Methodologies reservoir-simulation model was built for
range of reservoir permeability studied Multiphase-flow study and performance saturated reservoirs and used composi-
or by the single-phase-flow models that comparison of transversely fractured tional reservoir-simulation methods and
honored the reservoir-fluid properties of
This article, written by Special Publications Editor Adam Wilson, contains highlights of Eagle Ford oil.
For each of the three reservoir-fluid
paper SPE 181813, “Multiphase-Flow Performance Comparison of Multiple-Fractured
types—dry gas, black oil, and composi-
Transverse Horizontal Wells vs. Longitudinal Wells in Tight and Unconventional
tional oil—72 static reservoir-simulation
Reservoirs With Stress-Dependent Permeability,” by Rashid S. Kassim, SPE, Missouri models were built to study the effect on
University of Science and Technology; Larry K. Britt, SPE, NSI Fracturing; and well performance of well azimuth, res-
Shari Dunn-Norman and Fen Yang, Missouri University of Science and Technology, ervoir permeability, and the number of
prepared for the 2016 SPE Asia Pacific Hydraulic Fracturing Conference, Beijing, fractures. This study investigated the ef-
24–26 August. The paper has not been peer reviewed. fect of stress-dependent permeability on
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
Principal Stresses
and Fracture Orientation
In a normal fault-stress environment, the
vertical stress is the primary principal
stress, the maximum horizontal stress is
the secondary principal stress, and the
minimum horizontal stress is least prin-
cipal stress. Fig. 1 shows a horizontal well
with 10 transverse fractures, which was
drilled in the direction of the minimum
horizontal stress. Fig. 2 shows a horizon-
tal well with four longitudinal fractures
that was drilled in the direction of the
maximum horizontal stress.
While this study compares the per-
formance of transversely fractured vs.
longitudinally fractured horizontal wells,
there is a well-documented uncertainty in able for low- and ultralow-permeability ervoir fluid used in the black-oil model
the direction and azimuth of the longitu- unconventional resources. was Permian Basin oil. There were two
dinal fracture configuration once it exits reasons for a black-oil reservoir simula-
the perforations. The direction of propa- Black-Oil-Type Reservoir (Three- tion to be selected and for Permian Basin
gation is dictated by the intermediate- Phase: Oil, Gas, and Water). The res- oil to be used. First, the Permian Basin
and far-field in-situ stresses. An earlier
study showed that deviation of more than
10° from the azimuth of the preferred
fracture plane would always create trans-
verse fractures as the fracture propagates
away from the wellbore.
e
Nr w simulator can model fairly without sig-
nificant errors. Second, test models were
run to compare the results of a black-
oil model with those of a compositional
tured vs. longitudinally fractured hori-
zontal wells using Eagle Ford oil in a
compositional reservoir simulator shows
three significant findings. First, less
fo
model using Permian Basin oil, and the oil but more gas was recovered (using
result showed that there was very little the Eagle Ford oil) compared with the
!
difference in recovery (of oil and gas). black-oil reservoir simulator, which used
Production Monitoring/Surveillance
Marc Kuck, SPE, Drilling and Completions Engineering Manager, Eni
Industry trends in production moni- The industry is becoming for drilling, fracturing, and comple-
toring and surveillance continue to tion design, along with their effect on
demonstrate significant input from the more efficient and production performance.
academic and research fields, practi- knowledgeable in its data The industry is becoming more effi-
cal applications, and new-technology cient and knowledgeable in its data gath-
introduction. Market conditions simi-
gathering, evaluation ering, evaluation methods, and oper-
lar to those of the previous year are methods, and operations. ations. We will be better equipped as
driving some of the same discussions industry activity begins to increase,
again this year. With the decline in new because of our ability to operate more
drilling, our peers are focusing on opti- experiencing a broadened application efficiently at current market conditions
mization of existing fields and exam- as a more-cost-effective alternative. or because of market improvements
ining past performance in prepara- The focus on existing assets also has or both meeting on middle ground,
tion for new future developments. In increased discussion on well integri- in the near future. The articles select-
many cases, these efforts are being per- ty and how this should be monitored ed for review demonstrate some of the
formed with fewer personnel and limit- and maintained. achievements of our peers from the
ed capital resources. As we look toward future improve- last year. JPT
Optimization of existing fields is ments, we are also taking advantage
inspiring new concepts for statisti- of the decline in new drilling activity
cal and predictive production model- to review performance from the past Recommended additional reading
ing, along with new views on decline several years. Specifically in regard at OnePetro: www.onepetro.org.
analysis. Virtual flow-measuring sys- to fracturing, we are investigating
tems, while not new technology, are the relationships between methods SPE 183463 Well-Productivity Analysis
Integrating Ultrahigh-Resolution LWD
Images and PLT: A Data-Integration Case
Study by H.M. Al Shehhi, ADCO, et al.
Marc Kuck, SPE, is a drilling and completions engineering man- SPE 179058 Well-Surveillance Operations
ager with Eni US Operating Company in Anchorage. He has and Data Analysis in Thermal Viscous-
worked with major operators and service companies in the oil and Crude Project in Kuwait—A Case Study
gas industry for more than 20 years, specializing in well opera- by Shaikha Al-Ballam, Kuwait Oil Company,
tions and completions technology. Kuck holds several patents et al.
and has authored multiple publications. He has worked in Alaska SPE 183535 Implementing Fiber-
for the past 14 years and holds a BS degree in mechanical engi- Optics Distributed Sensing as a Key
neering from the Colorado School of Mines. Kuck is a member of Surveillance Tool by S. Al Shoaibi,
the JPT Editorial Committee. Petroleum Development, et al.
A real-time production-surveillance
and -optimization system has
been developed to integrate available
developed techniques for continuous
model tuning. First, a complete inte-
grated production model (IPM) of the
ever, conventional VFM implementa-
tions treat each well separately and only
use a limited number of sensors for any
surveillance data with the objective production system was built, spanning given well. In addition, many current
of driving routine production reservoir inflow and wells on the one VFM implementations require inputs
optimization. The system aims to hand and flowlines, risers, and topsides such as water cut and gas/oil ratio (GOR)
streamline data capture, automate on the other. The IPM is embedded in of the well from the last well test. The
data quality assurance, integrate high- a field management software platform flow-rate estimates therefore will be bi-
and low-frequency data to extract that houses several standard work flows ased as these values change over time.
maximum value, optimize the design as well as proprietary algorithms. Calcu- To overcome these issues, the chose
and analysis of commingled well tests, lation results and real-time data are vi- software provides a multiwell alloca-
and provide real-time multiphase sualized with a visualization application. tion (MWA) work flow. Although MWA
well-rate estimates for continuous As with other digital-oilfield ap- can close the overall material balance
well-performance evaluation. plications, the key is continuous cali- of the field, it relies strongly on the ac-
bration of the physical models when curacy of the wellhead pressure (WHP)
Introduction field producing conditions change. The for each well.
The technology was piloted in an off- standard solution within the chosen In cases where many of the WHP
shore field consisting of stacked deep- field-management software relies on sensors are either broken or have
water channel deposits developed with calibrating the well models with single- drifted, and where some wells have
five individual subsea drill centers, 15 producing-well tests. However, most of no working sensors at all, the VFM
active oil producers, one gas injector, the subsea wells in the asset have lost work flows provided by the cho-
and five water injectors. Equipment the ability to flow alone, and only com- sen software cannot be used direct-
is controlled remotely, and produced mingled well tests are available. Opera- ly. To address this issue, a customized
fluids are routed to surface by two 8-in. tors thus resort to a test-by-difference daily-well-rate-estimation (DWRE)
risers from each drill center. Produced protocol to obtain periodic well rates work flow was developed, which, like
gas is treated for use as fuel gas on for allocation, which introduces sig- MWA, also expresses the well-rate-
the surface and for gas lift by reinject- nificant errors when test conditions estimation problem as a nonlinear least-
ing into each production well. All wells deviate largely from production con- squares problem in which the produc-
were originally deployed with a full suite ditions. To overcome this issue, a cus- tivity index (PI), water cut, and GOR
of instrumentation, as well as valve- tomized commingled well-test-analysis are unknowns for each well. However,
status sensors for each well flowline (CWTA) work flow was developed to instead of using individual-well perfor-
and riser. The three-phase separators on analyze multisegment commingled well mance models to predict sensor mea-
the surface use orifice plate meters for tests simultaneously. surements as in MWA, a network model
rate measurement. Calibrated models coupled with real- is used and the reservoir pressures and
The approach taken was to integrate time field sensor measurements allow separator pressures are set as bound-
real-time data and physical models for for virtual flowmetering (VFM). Well ary conditions. This reduces the strong
real-time production surveillance and rates can be calculated in multiple ways dependence on individual-sensor data
optimization. The deployed system uses using the same sensor data but with dif- such that the work flow continues func-
software that incorporates in-house- ferent well-performance models. How- tioning even if sensors fail.
Implementation Overview
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of
The deployed real-time production-
paper SPE 181103, “Real-Time Production Surveillance and Optimization at a Mature
surveillance and -optimization sys-
Subsea Asset,” by Xiang Ma, Zachary Borden, Paul Porto, Damian Burch, Nancy tem consists of the standalone CWTA
Huang, Paul Benkendorfer, Lynne Bouquet, Peng Xu, Cassandra Swanberg, module and a software platform with
Lynne Hoefer, Daniel F. Barber, and Tom C. Ryan, ExxonMobil, prepared for the DWRE. After commingled well tests
2016 SPE Intelligent Energy Conference and Exhibition, Aberdeen, 6–8 September. are performed, the engineer uses
The paper has not been peer reviewed. CWTA to analyze the captured data,
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
Fig. 1—Four examples of petrophysical/geological parameters projected on a grid along with 365-day BOE/ft bubbles:
(a) water-cut grid inferred from actual well production; (b) HPV grid; (c) average resistivity×net pay (Ω•m-ft);
(d) aeromagnetic anomaly. All data are for Niobrara only, except aeromagnetic data, which is not formation-specific.
of describing the variance in production, the dependent and independent fracturing cost as a function of bar-
which variables are most important, and variables. rels of fluid pumped. This informa-
the relative influence of those parame- ◗ Run statistical diagnostics. tion can simply be stripped from a
ters on the production response. To this ◗ Check residual plots to ensure that fracture-treatment-cost proposal.
end, each mapped variable was sampled regression assumptions are not Sensitivity to a few parameters that
at a resolution of 1 sq mile (section level) violated. Also examine the plot of sometimes are produced by the multi-
and assigned to any producing horizon- predicted vs. observed response. variate analysis is harder to obtain and
tal well with a surface location within ◗ Assess the candidate-model- requires multiple cost proposals using a
that section. prediction performance by running fixed proppant volume.
sensitivities. Ultimately, these fracturing-cost
Database Maintenance The goal of a typical multivariate changes are compared relative to the
and Limitations analysis is to understand as much about change in well cost, or the numerator
The current DJ database contains com- the reservoir and stimulation as possible in the USD/BOE ratio. The approximate
pletion, petrophysical, and production and how each variable contributes to the well cost for different operators in the DJ
data for nearly 4,000 horizontal wells. overall success of a well. Basin is generally available through in-
For every well, this database contains Finally, what the reader needs to un- vestor presentations.
18 production parameters, 37 geologi- derstand is that this is a “dumb” analysis. The authors are interested in lowering
cal and reservoir-engineering parame- A multivariate-analysis model could pro- USD/BOE and in determining the effect
ters, and 59 completion parameters. The duce potential correlations between pa- of a completion change in terms of how
authors recognize that this production rameters, but it follows no physical logic it changes the total well cost and produc-
database is lacking in a few important re- to determine these correlations. A poten- tion history.
spects (missing completion parameters, tial problem associated with the lack of A main assumption in the cost analysis
incomplete geological parameters, and experimentation for DJ Basin fracturing is that the uplifts observed in production
production interference), but it also has designs is that the multivariate analysis [within the time frame of the multivari-
significant strengths. For one, it is a very does not excel at predicting beyond the ate analysis (i.e., 365 producing days in
quick tool to conduct comparisons and range of the data set; in other words, it this case)] are also the uplifts in the es-
a scoping analysis of trends to obtain an cannot predict what has not been done. timated ultimate recovery (EUR) of the
initial understanding of what appears to The authors recommend that, at the end well. Therefore, if a 10% uplift in 365-day
work in a specific area. It is very easy to of the scoping analysis, a cross check cumulative production is observed, it can
filter and compare wells by location, op- be conducted with a “smart” calibrated be assumed that 10% uplift also applies
erator, formation, vintage, and a range in fracture-growth model tied to a reser- to the EUR of the well.
completion types. voir model. When incremental cost is compared
with incremental revenue after 365 pro-
Brief Multivariate-Analysis Simple Well-Cost Model ducing days for Niobrara and Codell wells,
Background To determine cost sensitivity to all com- it is seen that suggested fracture-design
The purpose of multiple regression, or pletion parameters that might come out changes aiming to increase produc-
more-general multivariate analysis, is to of the multivariate analysis, the authors tion by 10% will reduce USD/BOE
determine the effect of completion and broke down the cost of a fracture treat- even on a 365-day production metric
petrophysical (geological) parameters ment by the parameters used in the multi- and, in most areas, at a net oil price as
on production. In general, the following variate analysis. This can be straightfor- low as USD 20/bbl.
steps are taken as part of an analysis: ward for some parameters. However, if Incremental revenue gains will far out-
◗ Define objective of study and the multivariate analysis uses lbm/ft, the strip the incremental costs—spending
quality-control data. cost needs to be calculated in USD/lbm/ft a few percent of well cost on the com-
◗ Perform exploratory data analysis by dividing the total proppant cost pletion improvements generally reduces
and check for outliers. Apply for a well by the proppant loading per USD/BOE by more than double the well-
transformation if necessary. Identify lateral foot. A similar logic applies to cost increase. JPT
◗
T his paper describes how the
use of production-modeling
tools, coupled with field pressure/
Kairi 2 (K2) WPP: nine producers
and three injectors
◗ Aripo (A) WPP: three producers
resentative well tests on all of the wells.
However, as the field entered its mature
phase, liquid rates of many wells fell out-
temperature-data-acquisition systems The production from the wells in each side the design range of the test-separator
and programming software, served as a WPP is commingled in each platform’s meters. Also, there were periods when is-
means to improve production allocation header and routed as follows: sues developed with the MPFM, affecting
and surveillance on a real-time basis ◗ Canteen 1 (C1) WPP: one subsea the accuracy of some well tests. Opera-
in the Greater Angostura Field. The production flowline to the low- tional constraints often affected the fre-
complete paper discusses a practical pressure gathering system quency of well tests available for use, and
approach to obtain accurate real-time ◗ Kairi (K1) WPP: two subsea operating conditions during well tests
well-production rates without the need production flowlines to either the may be different compared with normal
for purchasing costly flowmeters for low-pressure or the high-pressure operations for some wells.
each well. It also demonstrates how gathering system The combination of these issues
virtual metering can be used to identify ◗ Kairi (K2) WPP: two bridge- began to affect the allocation factor (AF)
production-optimization opportunities connected flowlines to either low- across the fields and reservoirs. AFs are
in the production system more readily. or high-pressure gathering system determined over time on the basis of a
◗ Aripo (A) WPP: one subsea comparison of the sums of test data of
Introduction production flowline to GEP oil, gas, and water production vs. the
The Greater Angostura Field is located 25 The oil produced from C1, K1, and K2 actual total production measured for fis-
miles offshore Trinidad and Tobago. The is processed on the CPP and pumped to cal and regulatory purposes. These com-
field was developed in three phases: Phase shore. The gas produced from these three parisons were normally conducted on a
1, which includes the Kairi, Canteen, and WPPs is distributed to export gas. The monthly basis.
Horsts fields; Phase 2, which includes production of the Aripo WPP is sent di-
the Aripo gas development; and Phase 3, rectly to the GEP for sales. Improved Methodology
which was scheduled to have been com- Accurate production allocation un-
pleted by the second half of 2016. Historical Well Testing derpins effective production and res-
At the time this paper was written, and Production Allocation ervoir surveillance, evaluation of in-
the Angostura development consisted of The historical allocation methodology fill and intervention opportunities, and
21 producers and seven gas injectors. was designed to allocate oil, gas, and production-optimization efforts. There-
The facilities and gathering system in- water production and gas injection to fore, in order to mitigate some of the un-
cluded one manned central-processing wells and reservoirs and to account for certainties associated with the allocation
platform (CPP), four unmanned well- well downtime. Wells are tested periodi- approach, a model-based methodology
protector platforms (WPPs), and one gas- cally in a test separator or a multiphase was developed to integrate periodic well
export platform (GEP). All wells have flowmeter (MPFM), depending upon the tests with ongoing gathering-system op-
dry production trees and are distributed well. The C1 wells can be tested only erating conditions.
as follows: through an MPFM, K1 wells through ei-
◗ Canteen 1 (C1) WPP: five producers ther an MPFM or a test separator, and K2 Production Model. The first step was
and two injectors wells only through a test separator. to build a robust production model
◗ Kairi 1 (K1) WPP: three producers The accuracy of this method of allo- capable of simulating the pressure/
and two injectors cation is dependent upon frequent, rep- temperature behavior of the wells and
gathering system, as well as honoring
the material balance for all three streams
This article, written by JPT Technology Editor Chris Carpenter, contains highlights
of production. The model was built with
of paper SPE 180887, “A Cost-Effective Methodology for Production Metering and a bottom-up approach from the sandface
Allocation Using Real-Time Virtual Metering in a Mature Offshore Oil Field: A Case to the separator.
Study of the Greater Angostura Field,” by Italo Raffaele Acuna, BHP Billiton, prepared Individual-well models were built to
for the 2016 SPE Trinidad and Tobago Section Energy Resources Conference, Port of reflect deliverability on the basis of in-
Spain, Trinidad and Tobago, 13–15 June. The paper has not been peer reviewed. flow and outflow relationships. To cap-
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
Heavy Oil
Tayfun Babadagli, SPE, Professor, University of Alberta
Over a 6-month time frame in 2016, I Low-interfacial-tension for additional reading and one experi-
was able to attend three SPE conferences mental work as a summary paper on this
on heavy oil in different countries span- (microemulsion) subject. Apparently, modeling efforts on
ning three continents (Canada, Peru, and and low-salinity injection advanced (but unconventional) technol-
Kuwait). Despite regional differences in ogies such as electromagnetic heating
the applications, potentials, problems,
in heavy oils in sands have continued. You will find a detailed
and technological needs, the common and carbonates mathematical analysis of the process in
theme in all conferences was “low cost.” and wettability alteration one of the papers summarized.
Cost optimization in heavy-oil produc- Despite the recent economic down-
tion was discussed from technical and in carbonates were common turn, we were able to hear the outcome
economic perspectives, not only in the topics at conferences of current field practices at pilot or dem-
technical sessions but also in numerous onstration scale. Papers detailing small-
panel discussions. held over the past year. scale cyclic-steam-injection applications
Such optimization can be achieved in Kuwait and Oman were worth reading,
through numerical modeling to suggest ing techniques as suggested reading in and one article on this specific subject is
general optimal strategies and develop- this issue. included here. Considering these activi-
ment plans or by using proper real-time Moreover, chemical and nanomate- ties in the Middle East, effective transfer
data acquisition (production optimiza- rial additives to water and steam have of technologies from North America to
tion) for prompt decisions while oper- received a great deal of attention. Low- that part of the world will become highly
ations are ongoing. This requires con- interfacial-tension (microemulsion) and critical in the near future. JPT
tinuous monitoring of the processes as low-salinity injection in heavy oils in
seen in many steam-assisted-gravity- sands and carbonates and wettabili-
drainage operations or other types of ty alteration in carbonates were com- Recommended additional reading
steam-injection applications. I selected mon topics at conferences held over the at OnePetro: www.onepetro.org.
two papers about advanced monitor- past year. I selected one review paper
SPE 181160 State-of-the-Art Review of
the Steam Foam Process by Eric Delamaide,
IFP Technologies Canada, et al.
Tayfun Babadagli, SPE, is a professor in the Civil and
SPE 180732 An Integrated Probabilistic
Environmental Engineering Department, School of Mining and
Work Flow for Primary and Thermal
Petroleum Engineering, at the University of Alberta, where he Performance Prediction of a Large
holds an NSERC-Industrial Research Chair in Unconventional Oil Extraheavy-Oil Field by Raushan Kumar,
Recovery. His areas of interest include modeling fluid and heat Chevron, et al.
flow in heterogeneous and fractured reservoirs, reservoir charac-
SPE 181431 Horizontal Steam-Injection
terization through stochastic and fractal methods, optimization Flow Profiling Using Fiber Optics
of oil/heavy-oil recovery by conventional/unconventional by Mahdy Shirdel, Chevron Energy
enhanced-oil-recovery methods, and carbon dioxide sequestration. Babadagli holds Technology Company, et al.
BS and MS degrees from Istanbul Technical University and MS and PhD degrees from
SPE 180726 SAGD Production
the University of Southern California, all in petroleum engineering. He was an execu- Observations Using Fiber-Optic
tive editor for SPE Reservoir Evaluation & Engineering during 2010–13 and is a mem- Distributed Acoustic and Temperature
ber of the JPT Editorial Committee. Babadagli received SPE A Peer Apart status in Sensing: SAGD DAS—Listening to Wells
2013, was elected an SPE Distinguished Member in 2013, and was an SPE Distinguished To Improve Understanding of Inflow
Lecturer in 2013–14. by Warren MacPhail, Devon, et al.
A. Injection
350
ranging from 200 to 400,000 cp.
Because of the high viscosity, first
300
production was considered possible B. Soaking
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
Materials
T he production of heavy and
extraheavy oil is difficult because
of its rheological properties caused by
fined sandpack in steam-injection sce-
narios in the absence and presence of a
water-based nanofluid. A displacement
Fumed silica nanoparticles functional-
ized with 1 wt% nickel and palladium ox-
high asphaltene content. Upgrading test was conducted by (1) construct- ides (SNi1Pd1) were used as a catalyst in
these unconventional oils requires ing the base curves, (2) estimating the the steamflooding process. Distilled water
large amounts of energy, decreasing oil recovery by continuous injection of and salt precursors of Ni(NO3)2•6H2O and
the production’s cost-effectiveness. vapor in the absence of the nanofluid, Pd(NO3)2 were used for nanoparticle func-
Nanoparticulated catalysts have and (3) identifying the influence of the tionalization. Deionized water was used
been shown to improve enhanced nanoparticles on the enhanced recov- for steam generation and for brine and
recovery of these oils by altering ery of oil. nanofluid preparation. A synthetic brine
their physicochemical properties, Results suggested that functional- of 2000 mg/L NaCl was also used for injec-
including asphaltene content. This ized nanoparticles lead to higher ad- tion into the porous media and for nano-
paper presents an investigation into sorption of asphaltenes, higher degrees fluid preparation.
the effect of catalytic nanoparticles of asphaltene self-association, and low-
on the efficiency of recovery from ered temperature of n-C7 asphaltene Porous Media. Before use, the prepared
continuous steam injection. steam gasification. Oil recovery in- porous bed was initially cleaned with a
creased to 46% for the system assisted 1:1 solution of toluene and methanol at
Introduction by nanoparticles in comparison with 10 mL per 5 g of sand. The sand then was
Several in-situ techniques have been that from vapor injection without the washed with deionized water and hydro-
used to enhance heavy- and extraheavy- nanoparticles. The °API value of the oil chloric acid to remove any dust or sur-
oil recovery with the objective of up- increased from 7.2 to 12.1 °API. A reduc- face impurities. Further, the sand was
grading the oil and improving its vis- tion of 59% in oil apparent viscosity placed in a vacuum oven at 120°C for 12
cosity and mobility. These techniques was also observed. Additionally, analy- hours to evaporate any remaining sol-
include thermal processes, such as sis of n-C7 asphaltene and residue con- vent. Approximately 100 g of the sand
steam-assisted gravity drainage, and tent demonstrated that nanoparticles was transferred to a stainless-steel slim-
cold techniques, such as treatments are an excellent alternative for heavy- tube. The absolute permeability of the
with diluents. Cold processes improve and extraheavy oil upgrading in pro- porous media was measured by injecting
the crude oil by dilution or destabiliza- cesses involving vapor injection. n-C7 brine after sand packing.
tion and deposition of asphaltene com- asphaltene content decreased by 5.2
ponents in the reservoir with solvents percentile points after vapor injection Preparation of Nanofluid. The nano-
that have a direct effect on viscosity. in the presence of nanoparticles com- fluid was prepared by mixing 1 wt%
This paper evaluates the use of nanopar- pared with the unaltered extraheavy oil. of polysorbate 80 with distilled water
ticulated catalysts in steamflooding. Also, residue content decreased 47%. and dispersing 500 mg/L of bimetallic
The nanoparticles were selected Improved oil recovery by nanoparticle nanoparticles. The mean size of the bi-
through batch-adsorption experiments. injection could be attributed to three metallic nanoparticles in the aqueous so-
A methodology was developed for eval- main reasons: (1) wettability altera- lution was approximately 61 nm, as mea-
uating the effect of the nanoparticu- tion of the porous media, (2) viscosity sured by dynamic light scattering.
lated catalyst in continuous vapor in- reduction because of reduction of the
jection. Oil recovery was evaluated by asphaltene aggregate, and (3) crude- Methods
use of a slimtube filled with a noncon- oil upgrading. Adsorption Experiments. Bath-mode
adsorption experiments were conducted
by adding a fixed amount of nanopar-
This article, written by Special Publications Editor Adam Wilson, contains highlights
ticles to the heavy-oil model solutions
of paper SPE 179699, “Heavy-Oil Upgrading and Enhanced Recovery in a Continuous and mixing. Then, nanoparticles with
Steam-Injection Process Assisted by Nanoparticulated Catalysts,” by C.A. Franco, adsorbed asphaltenes were separat-
L. Cardona, S.H. Lopera, J.M. Mejía, and F.B. Cortés, National University of ed from the mixture by centrifugation.
Colombia at Medellín, prepared for the 2016 SPE Improved Oil Recovery Conference, After that, the nanoparticles contain-
Tulsa, 11–13 April. The paper has not been peer reviewed. ing adsorbed asphaltenes were dried to
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
Recovery (%)
Thermogravimetric Analysis of n-C7 60
Asphaltenes. In order to evaluate
the catalytic behavior of the select- 40
ed nanoparticles, catalytic steam gas-
ification of adsorbed asphaltenes over
20
nanoparticles was conducted and stud-
ied by use of simultaneous thermogravi-
metric analysis. 0
0 10,000 20,000 30,000 40,000 50,000
Fluid-Injection Test. Nanofluid mix- Vapor PVI
tures were injected into the porous
media with a positive-displacement Vapor after nanofluid Vapor
pump. A leak test was performed by
pressurizing the packed-bed reactor. Fig. 1—Oil-recovery curves for vapor injection with and without nanoparticles.
The main objective of this displacement
test was to evaluate the effectiveness of
the nanoparticles in changing the physi- face. Hence, a heterogeneous adsor- tinuous vapor injection with and with-
cochemical properties of the extraheavy bent with multiple selectivity toward as- out nanoparticle assistance. The oil re-
oil and in enhancing the effectiveness phaltenes would be more efficient than covery from vapor injection without
of continuous vapor injection. The dis- that with single selectivity. nanoparticles stopped after approxi-
placement test was conducted by (1) mately 12,000 pore volumes injected
constructing the base curves, (2) es- Catalytic Steam Gasification of n-C7 (PVI) of vapor, reaching a recovery of
timating the oil recovery by continu- Asphaltenes. For catalysis to occur, the 34%. After 43,000 PVI, the recovery
ous injection of vapor in the absence asphaltenes first must migrate from the was 36%. Oil recovery is enhanced by
of nanofluid, and (3) identifying the in- bulk phase to the nanoparticles. The ve- vapor injection because of volatilization
fluence of the nanoparticles on the en- locity with which asphaltenes are ad- of light hydrocarbons, viscosity reduc-
hanced recovery of oil. sorbed on the active sites would depend tion, thermal expansion, variation of
on the adsorption potential of the se- relative permeability and capillary pres-
Analytical Methods. Changes in grav- lected nanoparticles. This is also de- sure, and gravitational segregation.
ity, oil viscosity, and n-C7 asphaltene pendent on their structure and chemi- Recovery using vapor with nanopar-
and residue content were determined to cal nature. While asphaltenes migrate ticles is enhanced over that of using
measure the effect of the nanoparticles to the adsorbent surface, a disaggrega- vapor without nanoparticles for sev-
on upgrading the extraheavy oil before tion of the asphaltene aggregate could eral reasons. Nanoparticles may alter
and after recovery with vapor injection. occur. Once asphaltenes are adsorbed, the system’s wettability toward being
they could be present in the active sites more water-wet. In addition to wetta-
Results and Discussion as monomers, dimers, i-mers, or aggre- bility changes, asphaltene adsorption
n-C7 Asphaltene Adsorption. A num- gates, depending on the uptake and the onto the injected nanoparticles also
ber of factors can influence the adsorp- crowding of the available active sites. leads to an increase in oil recovery. Once
tion of asphaltene onto nanoparticles, Then, as the temperature increases, the nanoparticles interact with crude oil, as-
including adsorbent surface chemis- adsorbed species would be cracked and phaltenes are increasingly attracted to
try, heterogeneity of adsorbent surface, free radicals will react with the H2 or being adsorbed. Because the asphaltene/
particle size, and asphaltene chemical O2 from the H2O(g) that was previous- asphaltene interactions are lower than
structure. Asphaltene adsorption was ly adsorbed and disseminated from the those of asphaltene/nanoparticles, as-
higher for SNi1Pd1 than for fumed sili- nanoparticle matrix to the active sites. It phaltenes would be removed from the
ca, indicating that supported nanoparti- is expected that the catalytic activity of aggregation system in the oil matrix
cles favored asphaltene adsorption. This the nanoparticles would affect the way and diffuse through the bulk phase to
could be because of the synergistic ef- both asphaltenes and water adsorb on the nanoparticle surface until maximum
fect of the compounds that enhances the the active sites, leading to a less expen- adsorption capacity is reached. As-
adsorption through multiple selectivity sive process in energetic terms. phaltenes remaining in the bulk phase
of the adsorbent. In addition, bimetal- would form smaller aggregates, and
lic compounds have synergistic effects Continuous Steam Injection Assisted reduction in the asphaltene aggregate
in avoiding sintering processes, which by Nanoparticulated Catalysts. Fig. 1 directly leads to a decrease in crude-
would lead to a less-heterogeneous sur- shows oil-recovery curves from con- oil viscosity. JPT
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
Seismic Applications
Mark S. Egan, SPE, Consulting Geophysicist
My career in the oil and gas indus- Indeed, this tighter sampled data that led to striking advanc-
try started in 1975. Shortly thereafter, es in imaging complex structures (espe-
a theme that became wildly popular in integration of seismic cially in subsalt settings) and provided
conferences, workshops, and internal with geomechanics the wherewithal for characterizing natu-
boardrooms was better integration of ral fractures and stress in both conven-
seismic with other oilfield disciplines.
energized the rise of the tional and unconventional plays. Indeed,
Initial progress was slow, but, if we fast- microseismic technique. this tighter integration of seismic with
forward to today, we see that significant geomechanics energized the rise of the
strides indeed have been made. A good microseismic technique.
representation of integration papers reservoir modeling. For example, in the Six integration papers were select-
appeared this past year, so that is the 1980s, a commonly used vibrator sweep ed. Three are presented here in sum-
theme we will use. in land surveys was 12–48 Hz. Today, mary format, and three are listed for
Advancement of the integration of sweeps start as low as 2 Hz and often additional reading. They come from all
seismic with other oilfield disciplines has go up to more than 100 Hz. The ben- over the world, cover a range of reser-
been fueled by two factors. First, evolu- efits to resolution and derisking have voir lithologies, and deal with a suite
tionary advances in seismic technologies been striking. of geologic settings. Let’s keep the
naturally led to more-powerful integra- Second, several developments in seis- progress moving. JPT
tion. A simple example is the progress mic were motivated in direct response
made in bandwidth. In this regard, to the needs of interpreters and engi-
advances in seismic sources, field record- neers. For instance, the appearance of Recommended additional reading
ing equipment, and data processing all creative, new, high-channel-count survey at OnePetro: www.onepetro.org.
benefited interpretation, inversion, and designs provided full-azimuth, densely
SPE 183289 Seismic Reservoir
Characterization of Carbonate Reservoirs—
Case Study of the Mishrif Formation,
Mark S. Egan, SPE, is a consulting geophysicist. He worked for Zubair Field, Iraq by J.M. Rodrigues, Eni,
Schlumberger and its heritage companies from 1975 to 2016, at et al.
which time he retired. Egan’s last position at Schlumberger was as IPTC 18956 Seismic-Driven Lithofacies
global chief area geophysicist in the Land Unconventionals Group for Geological-Model Infilling in a Deep
within the WesternGeco segment. His previous postings included Offshore Turbiditic Field—Case Study
chief geophysicist positions in North America, Saudi Arabia, by Victoriano Da Silva, Total E&P, et al.
Dubai, and London. Egan holds a PhD degree in geophysics, an
SPE 181872 Microseismic Mapping
MS degree in acoustics, and a BS degree in physics and mathe- Improves Understanding of a Complex
matics. He is a member of SPE, the Society of Exploration Geophysicists, the European Reservoir: A Case Study in a Southern
Association of Geoscientists and Engineers, and various local societies. Egan is a mem- Sichuan Shale Gas Field by Yang Wang,
ber of the JPT Editorial Committee and can be reached at egan9@hotmail.com. Halliburton, et al.
Qualitative 3D CTL
T he paper discusses an approach
for predicting the lateral variation
of net to gross (NTG) by use of 3D
ed from the seismic data, where the
static model predicts the presence of
the reservoir.
The objective of the qualitative 3D CTL
is to find obvious mismatches between
probabilistic seismic inversion. The the synthetic data generated from
goal is to define and understand the Field the model and the actual seismic. The
distribution of sands and shales on the The field is located in the Carnarvon sources of these mismatches have to be
basis of seismic reflection data. The Basin. The discovery was appraised by identified, and first-pass ideas for pos-
modeling and inversion are supported one well, which penetrated a thin sand sible corrections to the model to obtain
by the good quality of seismic data. layer of interdistributary bay reservoir a better match need to be made. Be-
This study underpins the benefits facies (referred to as Sand 1) and a thick- cause this is merely a qualitative check
of seismically constrained reservoir er distributary channel facies (referred before inversion, the loop is not yet
modeling. The use of probabilistic to as Sand 2). The overburden comprises closed with any model updates; this will
inversion to map geological features basinal shales and marls. be achieved later in the quantitative
is a new insight in the applicability The main steps in the 3D close-the- CTL step by use of model-based seis-
of this methodology. loop work flow are mic inversion.
1. The rock-property-trend models A static model with an initial 3D NTG
Introduction are derived using the available grid was available for the field. The first-
The study field is located in the Carnar- vertical wells in or near the field. pass static model designed to estimate
von Basin offshore western Australia. A “rock model” is a set of equations the gas initially in place is a simple
The field was appraised with one well, that honors well data and links constant-average-properties model de-
which has added incremental volumes the petrophysical properties such rived from the well. It is expected that
to existing nearby discoveries. Seis- as NTG and porosity with rock the seismic will give better control on
mically constrained reservoir-model acoustic properties. the properties away from the well. The
building creates models that are con- 2. A qualitative 3D check-the-loop regional rock models were used to pre-
strained by, and matched to, geolog- (CTL) step is carried out. This dict the elastic properties in the res-
ical concepts and seismic amplitudes involves the initial fit of the ervoir zone, in the overburden, and in
and travel times. The complete paper static model with the seismic the underburden shales. Fluid models
shows an approach using 3D probabi- data by forward modeling the for each fluid type (gas and water) were
listic seismic inversion to predict the synthetic seismic computed from built with the input from the pressure/
lateral variation in NTG over a proven the initial static model before volume/temperature report. Forward-
discovery. This, in turn, can be used to inversion. The quality-control modeled prior synthetics were generat-
update the static-model properties, im- step that follows aims to detect ed with a deterministic wavelet (derived
proving on the first-pass static-model mismatches between seismic and from the seismic) and compared with
build. Additionally, the inversion will synthetic data. the given seismic.
aid in the understanding of some poor- 3. Finally, constrain the reservoir
ly understood seismic expressions, es- model by use of the seismic data Prior Synthetic vs. Seismic. Fig. 1
pecially in the south of the field. In this by carrying out 3D probabilistic shows the comparison of the root-mean-
area, the authors noticed dim ampli- inversion of the key uncertain square (RMS) amplitude maps extracted
tudes on the amplitude maps extract- rock property. at Top Sand 2 reservoir (in a window of
−5 to +10 milliseconds) for the near
and far stack seismic and near and far
This article, written by JPT Technology Editor Chris Carpenter, contains highlights
prior synthetics, respectively. The RMS
of paper IPTC 18913, “3D Close the Loop Using Probabilistic Seismic Inversion
amplitude levels of the two maps show
for a Gas Field in the Carnarvon Basin, Australia,” by Shilpi Srivastava, Jeroen significant differences. Hence, there is
Goudswaard, Sito Busman, and Justin Ugbo, Shell, prepared for the 2016 a need to update the reservoir model
International Petroleum Technology Conference, Bangkok, Thailand, 14–16 November. properties, mainly the NTG. This will be
The paper has not been peer reviewed. Copyright 2016 International Petroleum achieved with a probabilistic seismic in-
Technology Conference. Reproduced by permission. version in order to close the loop.
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
7780000 7787119
7780000 7787119
1300
field. It is crucial to model the over-
1300
1219 1219 burden correctly such that it does not
1138 1138
1056 1056 influence the NTG estimation in the
975 975 channel sands below. Promise inver-
894 894
812 812 sion was carried out on multiple sce-
731 731 narios in order to model and capture
650 650
569 569 the overburden response correctly and
7770000
7770000
7759957
7759957
7780000 7787119
7780000 7787119
1300 1300
1219 1219 thetic match, as shown in Fig. 1. The
1138 1138 average NTG produced by the probabi-
1056 1056
975 975 listic inversion is used to update the res-
894 894
812 812 ervoir model and finally close the loop.
731 731 The results of the probabilistic inversion
650 650
do provide a number of model realiza-
7770000
7770000
569 569
488 488 tions; the standard deviations of these
406 406
325 325 models can be used as a measure to
244 244 build confidence regarding the inverted
7759957
7759957
162 162
81 81 results and how well the uncertain pa-
0 0
rameter has been constrained by seis-
Far Stack Seismic Prior Far Synthetic mic inversion.
Fig. 1—Comparison of RMS amplitude map between the near and far stack
seismic and prior near and far synthetic at Top Sand 2.
3D Inversion Results
The inversion results clearly showed the
distribution of good sands and reduced
3D Probabilistic it iteratively until a realization is found the reserves in both north and south,
Seismic Inversion that describes a good fit with the given immediately affecting developmental
The qualitative analysis described in seismic. The forward-modeled poste- decisions. Additionally, a channelized
the complete paper highlights some rior synthetics need to fit the actual feature can be recognized from the in-
key differences between modeled and seismic character well enough within version, which could not be interpret-
measured data in terms of amplitude the specified signal/noise ratio as de- ed from seismic amplitudes alone. This
differences. The current model pre- fined. Any model that does not satisfy helped in additional volume bookings.
dicts high amplitudes in the south of the specified noise level is rejected, and
the fault block, which is not support- a new realization is tested. The average Conclusions
ed by the seismic amplitudes. In the uncertain property from the final mod- The use of probabilistic inversion to
next step, the reservoir model is con- els can be accepted as the most like- delineate geological features such as
strained with seismic data by carrying ly solution that gives the best fit with higher-NTG channels is a new insight
out a 3D probabilistic seismic inversion the seismic. in the use of this methodology. It is
for NTG with a proprietary stochas- The acoustic properties in the input accomplished by using the probabilis-
tic model-based elastic inversion algo- static model are derived using the rock- tic inversion to predict the main static
rithm. This technology directly inverts property trends. Regional rock models reservoir property (NTG) in a geologi-
for static-model parameters, which were available that related the p-velocity cally constrained manner and then up-
sets it apart from traditional inversions and density to the porosity and shear dating the existing static model with
for impedances. velocity to p-velocity in the field. The this information. The results of the
The inversion takes a geological main property for uncertainty was NTG probabilistic inversion helped to in-
model with the prior reservoir proper- in Sand 1 and Sand 2 of the reservoir. terpret an additional channelized fea-
ties (NTG or porosity) and the given un- In the current static model, the over- ture that was included in the volumet-
certainties for a specific rock with a set burden comprises a wedge comprised ric estimation and helped in additional
of defined fluid properties and perturbs of marl, which shows significant thin- volume bookings. JPT
T hree-dimensional reservoir
models are best created with a
combination of well logs and 3D-seismic
conversion, the high-resolution 3D data
from seismic stochastic inversion and
well-log data were used as a direct input.
if they satisfy the criteria in the
process. These criteria measure
the degradation percentage of
data. However, the effective integration the synthetic trace related to
of these results is not easy because of Target Processing. In the first step, a the original trace. Thus, even
limited seismic resolution. With the wavelet-transform method was applied if degraded, some traces that
increasing quality of seismic data and to achieve noise elimination and resolu- somehow meet the criteria will
wide application of new methods, high- tion improvement. The principle of this be accepted.
resolution seismic-stochastic-inversion method includes the Hilbert spectrum 7. If accepted, the trace is considered
volume was used as a direct input to for nonlinear and nonstationary time- to be the solution to the inversion,
reduce the uncertainty of the reservoir series analysis. and it is incorporated into data;
model. Used as a direct input for then, a new trace is visited and
reservoir modeling, this method reduces Seismic Stochastic Inversion. The al- simulated.
the uncertainty of the model greatly. gorithm of seismic stochastic inver- 8. If the trace is not accepted, the
sion used in this study has the follow- same trace is resimulated until it
Introduction ing stages: satisfies the acceptance criteria.
A common method of reservoir model- 1. Initially, a sequential Gaussian The inversion process continues
ing is the stochastic modeling method. simulation is performed on the well until all the points of the grid are
The deficiency of this method is that it data of acoustic impedance, aimed simulated and all the resulting
can be difficult to determine the changes at filling the volume to be inverted. realizations are consistent with the
between wells. In order to describe the 2. The vertical and horizontal seismic data and with well data,
reservoir changes between wells better, variograms are generated and in addition to having minimum
sedimentary facies and 2D-seismic attri- modeled. residual with the seismic data.
bute tendencies were used as constraints 3. Each knot of the grid is revisited
in the process of modeling. However, be- randomly and resimulated. Time/Depth Conversion and Modeling.
cause every location has only one tenden- 4. For each knot, a synthetic trace is The algorithm of time/depth conversion
cy value, it is difficult to describe the res- calculated; for this, knowledge of in this study has the following stages:
ervoir overlay. the wavelet is needed. 1. Velocity modeling on the basis of
5. The synthetic trace is generated and the time/depth-relationship curve of
Method compared with the original trace, wells and seismic velocity data.
The work flows of this method, described where the residuals are calculated 2. Time/depth conversion for seismic
in greater detail in the complete paper, on the basis of the squares sum of stochastic inversion volume on the
essentially include three steps. The first the difference between the original basis of the velocity model.
step is target processing. Wavelet trans- trace and the synthetic trace. 3. Reservoir modeling on the basis of
form is applied to achieve noise elimina- 6. These residuals are then submitted depth seismic stochastic inversion
tion and resolution improvement. On the to an acceptance/rejection test. For volume and well data.
basis of the high-resolution seismic data, this, the optimization process by
the second step is seismic stochastic in- means of simulation annealing is Application
version. After the process of time/depth used. The residuals are accepted This method was used in H oil field in the
Middle East. H oil field is located in the
south of Mesopotamia Basin. The target,
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of
Formation N, was a structural-lithology
paper OTC 26404, “High-Resolution Seismic Stochastic Inversion as a Direct Input for
complex reservoir that was developed in
Reservoir Modeling,” by Chen Xin, Wei Xiao-Dong, Li Yan-Jing, Cui Yi, Ma Yingzhe, a long-axis anticline. The sedimentary
Yan Xiao-Huan, and Xia Yaliang, CNPC, and Wang Guan and Wang Xiaotian, environment is that of a barrier island;
China University of Petroleum, prepared for the 2016 Offshore Technology Conference the complicated reservoir is controlled
Asia, Kuala Lumpur, 22–25 March. The paper has not been peer reviewed. Copyright by barrier-island migration and tidal-
2016 Offshore Technology Conference. Reproduced by permission. channel reconstitution. The formation
3208
top-N N059
N060
3265
base-N
New-Porosity-Model Result Based on 3D High-Resolution Seismic Stochastic Inversion and Well Logs
3152 Well 1 New Well 02 Well 2
3208
top-N N059
N060
3265
base-N
Fig. 1—New Well 2 confirmed that the new model, which is based on 3D high-resolution seismic stochastic inversion and
well logs, is more reliable. Black well-log curve is GR.
thickness is 53 to 65 m, and the single- nonreservoir intervals, while the densi- the high-resolution seismic data, the pos-
sand-body thickness is 3 to 13 m. ty log can roughly distinguish reservoir sible number of solutions is reduced, thus
and nonreservoir, but cannot distinguish decreasing the nonuniqueness of the so-
Target Processing for Formation N. To good reservoir and poor reservoir. Wave lution. The seismic-stochastic-inversion
improve the ability of seismic data for res- impedance can better distinguish reser- results were multiple 3D volumes with the
ervoir identification, a wavelet-transform voir, poor reservoir, and nonreservoir. same vertical resolution as well-logging
method was applied to target processing Given such reservoir petrophysical data and good lateral-variation regularity
for Formation N. After the target process- features, to obtain practical inversion according to seismic attributes.
ing, the seismic response of the reser- results, seismic impedance inversion
voir sand in the high-resolution-seismic can be selected to invert the overall dis- Time/Depth Conversion and Model-
section becomes clearer than the seis- tribution trend because the formation ing for Formation N. To obtain accurate
mic data before target processing, and it thickness is approximately 60 m, which time/depth relationships, 40 wells were
matched the well-log interpretation. can achieve seismic-identification re- calibrated again by seismic-stochastic-
quirements. For the sand-body distri- inversion data. Then, velocity modeling
Seismic Stochastic Inversion for For- bution features of sublayers, under con- was established by accurate time/depth-
mation N. Good reservoir is mainly de- straints of seismic impedance inversion, relationship-curve and seismic-velocity
veloped as moderate- to fine-grained the GR-based seismic stochastic inver- data. Seismic-stochastic-inversion vol-
sandstone, poor reservoir is mainly de- sion is obtained to portray the reser- ume was converted from the time do-
veloped as argillaceous sandstone, and voir accurately. First, the probability- main to the depth domain on the basis
nonreservoir is mainly developed as density statistics and analysis of the GR of velocity modeling. Finally, reservoir
mudstone and sandy mudstone. Three curve and wave-impedance data are per- modeling was performed on the basis of
types of petrophysical analysis corre- formed, followed by the variogram es- depth seismic-stochastic-inversion vol-
sponding to these categories are dis- timation for the target interval. Then, ume and well data. The reservoir model
cussed in this study. seismic stochastic inversion can be based on the high-resolution 3D data
The analysis reveals that gamma ray carried out. from seismic stochastic inversion and
(GR) can distinguish lithologies well, be- Because the seismic-stochastic- well-log data will reduce the uncertain-
cause the reservoir is mainly the sand- inversion process is controlled not only ty greatly, especially in the area of very
stone with low GR (less than 45). The by the acoustic impedance features, vario- strong reservoir heterogeneity. A post-
acoustic log can distinguish reservoir and graphic model, and histogram but also by mortem is presented in Fig. 1. JPT
For a limited time, the complete paper is free to SPE members at www.spe.org/jpt.
Fig. 1—(a) Map of the two-well pad, with wells 600 ft apart with seismic-built
DFN calibrated with offset borehole-image logs. (b) Map of microseismic data
with DFN overlain in background.
properties were upscaled to the 3D relation along well trajectories was also
geological model and propagated with carried out.
geological surfaces. After the propaga- After the blind tests against four wells
tion, elastic and stress properties were from different parts of the seismic sur-
calculated, accounting for structur- vey, the spectral-edge-based SDP was
al variations. confirmed to be the best-matching seis-
mic representation of discrete fractures
DFN Model. The depth-converted seis- in the subsurface. Then, the SDP set was
mic data were used for generating three classified as an extracted fracture net-
SPE International different seismic discontinuity attri- work and its overall population char-
Conference on butes: most-negative curvature, spec- acteristics were studied, and the 3D
tral edge, and orthogonal decomposi- fracture network was converted into a
tion. These attributes were enhanced standard DFN model.
Oilfield further using the ant-track attribute.
In the next stage, seismic discontinuity
planes (SDPs) from all three attribute
By use of the extracted seismic-
scale fracture sets and their statisti-
cal properties, the subseismic por-
Chemistry cubes were extracted and overall popu-
lation statistics were assessed.
tions of each of those fracture sets
were modeled within a selected area,
Among the three SDP sets, the where hydraulic-fracture data and the
3–5 April 2017 curvature-based SDP set was not used geomechanical model are available for
Conference Center at La Torretta further, because it showed poor orien- validation of the survey. Orientations,
Montgomery, Texas tation statistics with respect to the dip density distributions, and length dis-
and azimuth of natural fractures known tributions were used directly to model
www.spe.org/go/networkoilfield
from the area. After removing SDPs that fractures with lengths ranging from 20
represent the acquisition footprints, to 1,000 ft.
local SDP dips from the orthogonal com-
Register early and ponent and spectral-edge-based ant- Hydraulic-Fracture Modeling and
save up to USD 300! track attributes were extracted along Calibration. The UFM was used
selected wells and plotted with the frac- in this study to model and calibrate
ture dips from borehole-image logs of the hydraulic-fracture profiles in the
those wells. In addition to the compar- Avalon Shale. Key inputs for proper
ison of overall orientation of SDP and hydraulic-fracture calibration are geo-
fracture dips per well, their detailed cor- mechanical model, a DFN, lamination
AVASTHI &
ASSOCIATES, INC.
Petroleum & Geological Engineers
9805 Katy Freeway, Suite 500
Reserve Reports
Fair Market Value
Estate Appraisals
Expert Witness
215 Union Blvd., Suite 350
Houston, TX 77024-1271
Lakewood, CO 80228-1840
Worldwide Energy Consulting, Ph: (713) 209-1100 Fax: (713) 209-1104 Telephone: 303/298-0860
Engineering, Training, Shale, CCUS, and e-mail: info@HuddlestonCo.com Facsimile: 303/298-0861
CO2 EOR/IOR Project Operations
Since 1990 SERVING THE OIL AND GAS INDUSTRY FOR
Field Development Planning (FDP) of Onshore and Offshore OVER 60 YEARS
(Shallow and Deep water) Oil, Gas, Gas-condensate, and
Heavy Oil Fields, and Unconventional/ Shale Resources.
Exploration • Development • Surveillance
Reservoir Characterization, Engineering and Simulation
(Static and Dynamic Modeling), Economic Evaluation, CO2 Enhanced Oil Recovery • Optimization MILLER AND LENTS, LTD.
EOR/IOR, CCUS, Geomechanics, and Fracturing/Stimulation
INTERNATIONAL OIL AND GAS CONSULTANTS
Global Head Office: 800 Rockmead Drive, Suite 212 Data-Driven
Houston, Texas 77339, U.S.A. • Phone: +1-281-359-2674 Reservoir Engineering
Specializing in All Phases of Reserves Evaluations,
Solutions
CG A &
Including Petroleum Economics,
Reservoir Engineering, Geology, and Petrophysics
• Field development studies
• Reservoir characterization
and simulation Two Houston Center Phone: (713) 651-9455
909 Fannin St., Ste. 1300 Fax: (713) 654-9914
• Environmental studies Houston, TX 77010 e-mail: mail@millerandlents.com
www.intera.com ipcinfo@intera.com Web pages: http://www.millerandlents.com
CAWLEY, GILLESPIE & ASSOCIATES, INC. Austin • (512) 425-2000 Houston • (281) 560-4560
PETROLEUM CONSULTANTS
International Reservoir
NITEC
SINCE 1960
FULL PHYSICS WITH NO SHORTCUTS SINGLE EXECUTABLE FOR WINDOWS & LINUX
Optional GPU acceleration is available, free of charge, in all new tNavigator releases,
starting March 2017. Blackoil, compositional, thermal compositional models are supported,
as well as every industry-standard input keyword formats. Specific acceleration factors
are model and hardware dependent. TNAVIGATOR.COM/GPU