P.
DONNELLY, AspenTech, Bedford, Massachusetts
Business Trends
EPC 2030: Five vital characteristics
that will define the EPC firm of tomorrow
The state of the global engineering, a brighter future are listed here. The fu- er startup, shutdown, turnup, turndown,
procurement and construction (EPC) ture’s leading EPC firms will be: steady-state, summer and winter cases to
industry can accurately be described as 1. Focused on design outcomes configure, size and rate plant equipment.
“challenging.” Operating in an environ- 2. Agile, high-velocity businesses By 2030, software and high-perfor-
ment of volatile, uncertain, complex and 3. Talent magnets mance computing will be used to consider
ambiguous conditions, EPC firms are fac- 4. Economically savvy at all levels of thousands of potential cases from a wider
ing unsustainable levels of stress. the organization range of ambient temperatures to more
From onerous contracting models that 5. Relentlessly collaborative. complex variations in the chemical make-
discourage collaboration, to unpredict- up of crude oil and other feedstocks.
able commodity prices that undermine Characteristic 1: Focused on design Leveraging machine-learning and high-
the attractiveness of the sector as an em- outcomes. The single biggest differen- performance computing, and setting up
ployer, the reasons are many and com- tiator for an EPC firm is the quality of its and running these cases to define global
plex. Combined, these stressors are the work. Quality manifests through lower optimums for design parameters, will be
forces that will reshape the industry over required capital expenditures (CAPEX), done quickly within a single interface—
the next decade. accurate and complete engineering data and without undue time spent on the
Where does that leave the innova- that helps the contractor avoid problems mechanics of comparing a vast number of
tors—the next generation of ambitious during construction, designs that meet potential cases and operating conditions.
engineers and executives who want to production targets while consuming less Getting there. The following hurdles
drive change and improve performance energy and emitting less pollution, and and recommendations include:
within their firms and across the indus- plants that safely exceed targeted uptime. • Hurdles to overcome: Available
try? The author’s company has identified Leading EPC firms in 2030 will play a technology for determining the
five characteristics that will define the critical role in designing the plants of the optimal number of cases needed
successful EPC firms of the future. These future by having technology, processes for a given set of desired outcomes
key traits will help companies effectively and cultural mindsets that make the de- and constraints, and for quickly
guide investment and spur innovation livery of high-quality work product their considering thousands of cases
throughout the digital transformation top priority. in a reasonable amount of time.
journey. Striving to embody these quali- While today’s designs are often “fin-
ties will drive positive changes in careers, ished” when the EPC firm runs out of
companies and the industry, making the time or budget, the designs of 2030 will
EPC segment of 2030 stronger, more in- be efficiently and rapidly optimized
novative and increasingly profitable. around a broader array of potential oper-
ating cases. This will be accomplished by
running thousands of simulations to find
FIVE CHARACTERICSTICS the most optimized configuration, size,
THAT WILL DEFINE THE EPC type and model of equipment that can
FIRMS OF THE FUTURE come together to operate as an efficient
According to conferences and various plant system (FIG. 1).
media, the digital revolution is largely Rapid design exploration. Depend-
underway. However, many EPC execu- ing on the complexity of a project, the en-
tives say that their companies have experi- gineering team may run several or a couple
mented with digital initiatives, but they dozen design cases—each one requiring FIG. 1. New technologies will enable the
are still deliberating on where they will engineering cycles and personnel hours— EPC firm of the future to run thousands
place their “big bets.” With that in mind, to fully understand the operating charac- of simulations to find the most optimized
the characteristics that can help shape teristics of a process plant. On the simple configuration, size, type and model of
how and where EPC firms will invest for end of the spectrum, engineers will consid- equipment to optimize plant operations.
Hydrocarbon Processing | JANUARY 2020 9
Business Trends
• Recommendations. EPC firms anecdotal experience, rather than bring together theoretical models
should invest in personnel and on using comprehensive data. with actual operating data to
technology to stay at the forefront • Recommendations: EPC firms troubleshoot equipment, optimize
of this looming advancement in must build probability-based feedrates, refine maintenance
process engineering. Software simulation into their early design intervals and avoid unanticipated
providers must invest in the processes. Software providers conditions and events.
applications and infrastructure should include decision-support
to facilitate multi-case engineering tools in their estimating and Characteristic 2: Agile, high-velocity
and high-performance computing. engineering software. businesses. According to a 2018 indus-
Probability analysis for early design Combining model types to im- try survey,1 two-thirds of EPC firms were
making—optioneering. A design team’s prove design and operations. The use undertaking, or planning to undertake, a
ability to impact final costs, in addition of first-principle models for defining and major initiative to improve their produc-
to the functional capabilities of a process predicting performance and outcomes is tivity, speed and agility. Some industry
plant, decreases rapidly with time. Early standard in the process industries. How- executives are targeting productivity im-
decisions—such as how many trains ever, some processes are more difficult to provements of at least 30% (measured in
will be used, how much buffer capac- predict, such as polymerization, catalyst work hours and dollars), and are striving
ity is needed, or how much redundancy degradation, bioprocesses and crystalliza- to be more responsive to customers, from
to build into critical components—lock tion. These processes are often managed bidding through project handover. The
in the design approach and limit sub- through less precise techniques, such as next section describes where those im-
sequent alternatives. Too often, these operator experience or rules of thumb, re- provements can be found.
critical early decisions are made based on sulting in sub-optimal performance. A shift from documents to data. In-
strong opinions, emotional positions or By 2030, leaders will embrace next- stead of engineering data being stored,
the experiences of one or more influen- generation capabilities to more accurately distributed and otherwise managed in
tial members of the design team. model complex, real-world behaviors. Hy- documents, it will be stored, managed
Also problematic is the speed at which brid models combining output from first- and accessed from multiple project da-
these decisions are made. Uncertainty principle models; operating models aug- tabases, on premise and/or in the cloud.
can slow down key decisions, holding mented by machine-learning; and black Data aggregation, rationalization and
up the project, impeding the progress of box models will be easier to build, main- cleansing will be done through a central
those relying on preceding steps to com- tain and deploy (FIG. 2). Actual operating enterprise data insights tool, with links to
plete their work, and causing the sched- data, along with data from lab experiments operating systems; design and engineer-
ule to slip. and pilot plants, will be leveraged in these ing systems; and corporate information
In 2030, EPC firms and owners will more precise models to improve predic- technology (IT) systems, such as enter-
utilize a more rigorous and data-driven tive power and better optimize everything prise resource planning (ERP) software.
approach to optioneering as part of their from catalyst replacement intervals in reac- Data will then be available to feed other
standard design processes. For example, tors to separation efficiencies in columns. systems or to inform decisions via cus-
thousands of Monte Carlo simulations Getting there. The following hurdles tomizable dashboards, with live links to
will be leveraged to simulate the results and recommendations include: the original sources of the data.
of alternative design scenarios to evaluate • Hurdles to overcome: Barriers and Data created during feasibility and
the likelihood of different outcomes, and lack of incentives for sharing data conceptual engineering will be shared
to quickly enable decisions that will im- between owners and EPC firms. with the front-end engineering design
prove cost, uptime and return on invest- • Recommendations: Owners and (FEED) contractor. Likewise, the FEED
ment (ROI) metrics. licensors can provide EPC firms data will be available electronically to the
Getting there. The following hurdles with samples of operating data, contractor performing detailed design
and recommendations include: lab data or data from pilot plants and construction. This digitization of
• Hurdles to overcome: that can be used to refine designs engineering data across disciplines and
Acquisition of historical data and to more precisely size and rate project phases will boost productivity by
to inform simulators; decision- equipment. During operations, avoiding rework, re-entry of data, errors,
making based on emotional and owners can leverage EPC firms to omissions, and time and effort required to
keep up with the inevitable changes, while
enabling faster turnaround times on bids,
design options and project deliverables.
Getting there. The following hurdles
and recommendations include:
• Hurdles to overcome: Contractual
and cultural norms, along with
misaligned incentives, that
discourage data-sharing; the use of
FIG. 2. By 2030, leaders will embrace next-generation capabilities to more accurately model documents as the primary vehicle
complex, real-world behaviors. for certifying engineering work.
10 JANUARY 2020 | HydrocarbonProcessing.com
Business Trends
• Recommendations: As part of major rationalization of their technol- partnerships that span workflows,
their digital journeys, EPC firms ogy platforms and providers. Chief in- functional areas and file formats,
should plan on digitizing and formation officers (CIOs) will shrink and that create links to adjacent
automating the flow of data across their portfolios to 20% of what they are systems, such as ERP.
their organizations and across most at present, and will make strategic in-
of their internal systems. They vestments in a few key suppliers that of- Characteristic 3: Talent magnets.
should also prepare to extend the fer broader solution coverage and make The oil and gas industry is at the forefront
sharing of data with external parties, it easier to move data around. This will of engineering, science and technology
such as suppliers and customers. dramatically lower costs, foster collabo- innovations on a scale that is dazzling,
Software and technology providers ration and speed up projects. albeit mostly hidden from the general
must work together to smooth Getting there. The following hurdles public. Despite the Herculean feats being
interfaces and the flow of data and recommendations include: accomplished daily, the energy and chem-
between engineering and business • Hurdles to overcome: icals industries have been on the receiving
software applications, as well as Customer requirements for end of a steady stream of negative media
to make their data more easily deliverability in specific file coverage that has affected the perception
available for data aggregation, formats, and incompatibility of and attractiveness of these sectors as em-
integration and dashboarding tools. software products and data formats. ployers. Coupled with the cyclical nature
Streamlining software and tech- • Recommendations: Engineering of commodity prices and the resultant im-
nology. It is not uncommon for the en- and IT departments should join pact on employment, the ability to attract
gineering function to buy, support and forces to catalog and assess each of and retain talent in both engineering and
maintain 100–200 different engineering their applications, and to develop in the trades has suffered.
software applications and technology a plan for portfolio rationalization, By 2030, EPC firms, driven by cost
platforms, most of which are distinct, with a goal of consolidating to the pressures, skilled labor shortages and the
incompatible islands of information. few core suppliers that provide need for productivity enhancements, will
This is unwieldy and expensive, and acts the largest value and portfolio. have automated many routine tasks. In
as a barrier to speed and collaboration. Software and technology suppliers addition, knowledge automation enabled
By 2030, leaders will have undertaken a must acquire technology and form by the combination of operational data,
What if...
... you could add full feedstock
flexibility to your refinery?
You can. We can give you the ability to turn practically any renewable feedstock
into an on-spec fuel. So, you can future-proof your refinery against changing
demands and take advantage of new opportunities to extend your market share.
HydroFlex™ - Fuel for thought
topsoe.com
Hydrocarbon Processing | JANUARY 2020 11
Business Trends
artificial intelligence (AI), domain exper- and data from operating plants. ect and portfolio decisions.
tise, guidelines, best practices, local codes • Recommendations: EPC firms Getting there. The following hurdles
and previous designs will help guide en- should invest in automating and recommendations include:
gineers to more optimal design decisions interfaces and eliminating manual • Hurdles to overcome: Internal
faster. This will free up engineers to focus data handoffs. Engineering software functional silos and over-reliance
on more compelling and higher-value en- providers must make their products on already constrained estimators.
gineering tasks, resulting in greater job easier to learn and use, enabling • Recommendations: Engineering
satisfaction and higher retention. newer engineers to get up to speed firms must invest in software that
As the industry pushes deeper into and be productive in months instead brings economic and engineering
finding new solutions for reducing en- of years. Owners need to recognize data together, while bringing their
ergy consumption, preventing unplanned the value of their operating data to estimating, engineering and project
emissions, and making the energy and inform future plant technology and controls teams closer together.
chemicals industry safer and more effi- designs, and to share more data with
cient, more young people will see the vast their EPC partners. Characteristic 5: Relentlessly col-
opportunities that these segments offer laborative. A tremendous amount of col-
for making meaningful and lasting im- Characteristic 4: Economically savvy laboration occurs on a typical project be-
pacts on society. at all levels of the organization. When tween technology licensors, multiple EPC
Automation knowledge work and an owner is evaluating the feasibility of a firms, engineers from many disciplines,
cognitive design guidance. Talented potential project, the focus is on develop- estimators, project managers, owners, sub-
and highly trained engineers are spending ing a sound economic assessment, albeit contractors, consultants and regulators.
too much time on mundane tasks, such as based on limited engineering data. As the The problem is that much of this collabo-
looking for information in multiple plac- project proceeds through conceptual and ration is happening through brute force
es, working from outdated information basic design, if the estimating and engi- and is rife with problems and limitations.
and manually entering data into multiple neering departments get out of sync, the Emailing spreadsheets that are soon out
systems—only to have to re-enter the result is a less reliable estimate and a po- of date, conference calls where key people
data when something changes. tential for unpleasant financial surprises are missing due to conflicts, and changes
By 2030, leaders will empower em- down the road. in scope or schedule that are not distribut-
ployees with more intelligent solutions By 2030, engineering and estimating ed in a timely manner result in thousands
that automate mundane and error-prone functions will become more integrated. of wasted work hours every year.
processes, and provide actionable insights EPC firms will leverage estimating sys- By 2030, leaders will break down silos
and guidance, leveraging embedded AI tems that are connected to engineering across the value chain to accelerate the ve-
and machine-learning, high-speed com- systems throughout the project, so that locity of their business. Technology will
puting, domain experience and historical the estimate keeps pace with design prog- have made inroads toward addressing and
data. Data will be centralized in fewer sys- ress and any changes in scope. Further- automating many of these pain points,
tems, and updates can be made centrally more, the estimate will be connected to but technology alone will not solve these
and then propagated across project partic- the project controls system, so the plan challenges. If not accompanied by chang-
ipants. Revision management will be stan- vs. actual data is readily available for use es in policy and mindset, initiatives aimed
dard, so that there is traceability and ac- in managing projects more efficiently to a at improvements will fail.
countability for engineering information, profitable and timely outcome. Business relationships evolve to
without having to rely on documents. Integrated project economics. In better align objectives and incentives.
Software enhanced with embedded 2030, leaders will incorporate better cost- In 2030, owners and contractors will have
AI will aid engineers by guiding them and risk-based decision-making into their evolved their approach to contracting to-
to more optimal decisions—and by le- operations, earlier in the process. Project ward fostering more open and collabora-
veraging design rules, historical and ex- controls systems will be fed directly from tive relationships, with better alignment
isting performance data, Monte Carlo the estimating system, and estimating across parties toward a good outcome
simulations, and firm-specific knowledge systems will get “smarter” by using actual for the project overall, not just individual
and standards. AI will alert engineers project data. Companies will share proj- participants. Incentives will be realigned
when they are deviating from previous ect data more broadly within their firms, to motivate all parties toward a set of
projects, sound engineering principles and rules and AI will assist with normaliz- common objectives that will drive suc-
or firm standards. Actual operating data ing historical data for use on new projects. cessful project outcomes.
from similar designs will be analyzed, and Integration of current and past project Owners will be incentivized to share
design changes will be suggested. engineering and economic data will be more data with their contractors—and
Getting there. The following hurdles available via customizable dashboards for contractors and subcontractors will be
and recommendations include: use throughout the firm. Estimators will willing to surface potential problems
• Hurdles to overcome: use this data for beginning new estimates, sooner, instead of trying to cover them
Engineering software that is a engineers will use the data for faster proj- up. Working together with aligned objec-
“system for experts,” rather than ect kickoffs, and executives will use the tives and incentives, the parties will find
being an “expert system”—and data to better manage projects to a profit- better solutions to problems faster, thus
access to previous design data able conclusion and to inform future proj- reducing the need for change orders and
12 JANUARY 2020 | HydrocarbonProcessing.com
