Hydrocarbon Processing - January 2022 - 18

Sustainability
2. Monitor key performance indicators (KPIs),
displaying data on user dashboards and identifying
problem areas.
3. Adjust dynamic targets based on plant states and
operating modes, while identifying abnormal
situations and likely process upsets.
4. Isolate root causes behind performance gaps and
provide suggested corrective actions.
5. Calculate setpoints to optimize energy use at
multiple operating levels, thereby minimizing cost.
The extent to which a given platform performs these functions
is the main predictor of overall project success. For example,
predictive models attached to dynamic targets can go
beyond simply identifying abnormal situations. A truly sophisticated
system should not only track these situations in real
time but also offer an analysis of what is going wrong in the
process (FIG. 2). In this way, the model can reach more deeply
into the interactions and offer practical ways to improve operation,
beyond just saving energy.
Saudi Aramco and EMIS. As a brief case in point, Saudi
Aramco launched an EMIS project at its Abqaiq facility, the
largest oil processing and crude stabilization plant in its fleet.
The installed EMIS covered 24 production units, 57 non-production
units and 237 pieces of critical equipment. Company
management wanted to improve on the facility's average annual
energy consumption of 30,000 Btus/bbl of oil equivalent.
In addition to many concerns common to most refiners, this
facility also had specific problems to solve, including:
* The existing energy monitoring systems reported
many false warnings of overconsumption that
were difficult to troubleshoot.
* The operators occasionally found themselves operating
to incorrect targets due to changes in the plant modes
and states driven by a shift to a different crude source.
* The operators were adjusting loads and feeds without
sufficient information to consider the full impacts on the
performance of each unit and its equipment.
Engineers designing the EMIS captured the right data to analyze
and solve the basic operational concerns, as well as those
unique to each unit. This data was combined with dashboards
(FIG. 3) configured to show critical variables, so operators
could maintain critical situational awareness of process operations
and energy use and then take appropriate action.
These actions were based on 13 optimization models, each
of which used nonlinear programming problems and mixed-integer
nonlinear programming problems methods to deliver results
for minimizing the energy consumption across all plants.
The results of the project included:
* An estimated energy cost savings of $22 MM
in the first year
* An increased operational performance visibility
for operators and management improved
decision making at all levels
* Faster problem-to-resolution cycle time
* The best setpoints and key parameters for highest
profitability retained and shared among operators.
PEMS: Extending capabilities of energy management.
Many combustion processes within a refinery are regulated by
the U.S. Environmental Protection Agency (EPA) or equivalent
in other countries. Such agencies require that most fired
heaters, boilers and similar equipment have a continuous
emissions monitoring system (CEMS) to measure and record
output of various air pollutants any time equipment is in operation.
Pollutants in this context can include sulfur dioxide,
nitrogen oxides, carbon monoxide, CO2
, ammonia, oxygen
and possibly others. The conventional solution in these cases
is to deploy one and potentially more gas analyzer technologies
designed to quantify the relevant pollutants for each source.
This approach can work well when applied correctly; howFIG.
2. An effective system does not just report data, it offers
suggestions for how operators can improve operations.
ever, it can become complex and expensive, particularly when
a given installation can burn multiple types of fuels capable of
producing different effluents. If there is an opportunity to save
money using an inexpensive fuel for some period, but the analyzer
cannot handle the likely effluents, it is a missed opportunity.
Moreover, given the number of likely applications in a
refinery, there can be dozens and perhaps hundreds of analyzer
installations, all requiring appropriate calibration, consumables
(specialized gases) and maintenance.
In many respects, the biggest drawback to this approach is
FIG. 3. This plant's system shows operators a variety of critical KPIs
so they have situational awareness of energy consumption and
how they can control it.
18 JANUARY 2022 | HydrocarbonProcessing.com
how little it helps improve the process or reduce costs. CEMS
analyzers are there for the sole purpose of monitoring emissions.
If operators in the control room can see what they are
reporting, which is usually not the case, they might get an indication
that something is wrong with the combustion, but this
is not always very helpful.
A better alternative is using a data-driven PEMS instead of
http://www.HydrocarbonProcessing.com
Hydrocarbon Processing - January 2022

Table of Contents for the Digital Edition of Hydrocarbon Processing - January 2022
