Best practice desalting

Part 1: crude flexibility window

A replacement programme of desalter treatment chemistry transformed the
operation of crude units struggling to meet performance targets

BRAD MASON
Nalco Champion

R
eliability remains king in
the refining environment 120
today, which can some- Brent
Oil price, $/BBL

100
times go against best laid plans WTI

when processing challenging 80 CDN light

WCS
crudes. This is the ongoing jug- 60
gling act: to capitalise on the 40
discounts presented by these
20
crudes while balancing all the
potential negative impacts; and
14

15

16

17
01

01

01
how to safely navigate these
20

20

20

20
l2

l2

l2
n

n
Ju

Ju

Ju
Ja

Ja

Ja

Ja
waters.
The articles that will fol-
low convey a vision for reach- Figure 1 Recent crude price history – WCS vs Brent/WTI/CDN Light Source:
ing a new level of performance www.oilsandsmagazine.com/energy-statistics/monthly-average-oil-prices-wti-brent-wcs
across the crude unit. The fol-
lowing articles will detail a ational-chemical (M-O-C) pro- The discussion will begin
‘crude unit of the future’ con- gramme approach, a refiner with two recent case studies
cept, whereby refiners will will never be able to capi- that reflect how critical desalt-
have the ability to make pro- talise on the potential ben- ing programmes can be for
cess changes to operations in efits this technology brings. solving the immediate process-
near-real time with the same Without stable operations in ing challenge such that further
crude blend, while also main- the desalter that control emul- optimisations can be made and
taining safety limits and key sion band growth and efflu- sustained.
performance indicators (KPI). ent quality, regardless of crude These sites run a challenging
This programme involves state slate processed, making pro- mix that often reaches 100%
of the art technologies from the cess changes to mix valve dP, of a Heavy Canadian crude
desalter, extended to the waste- level set point, mud wash blend, the main reason for
water treatment plant, and practices and so on will be this being potential improve-
across the crude unit overhead. unlikely. However, when you ment in operating margin. As
Before that, however, this can provide a ‘tried and true’ Figure 1 shows, the average
article will focus on best prac- best practice desalting pro- discount on the WCS bench-
tice desalting as a mandatory gramme that addresses indus- mark compared to a light bas-
requirement. Without a solid, try concerns, only then can you ket of crudes is still $10-15/
best practice mechanical-oper- optimise the process. bbl, and there are other Heavy

www.digitalrefining.com/article/1001414 PTQ Q3 2017 1

 treatment plant, but oil under-
Case study 1 desalter operations overview
carry remained excessive, and
became worse with the cost
Design charge rate, bpd 150 000 [historical ~110-115 000] advantaged crude slate the
Crude blend gravity, API 21-22° refinery needed to process. Oil
Design wash water rate, % of crude charge 5.0-8.0% [actual only 3.5-4.0]
Desalter design and configuration Petreco Bilectric; two-stage in series under-carry was such a con-
Residence time, minutes Oil = 6-8; water = 100-110 sistent problem that approx-
imately 30% of crude charge
Table 1 was bypassed around the first
stage of a double stage desalter.
Base Incumbent NC Incumbent Nalco When Nalco Champion was
case trial trial trial II Champion invited to trial, the immediate
3000
goal was to show that the treat-
grease, ppm
Oil and

2000
ment programme proposed
1000
– based on comprehensive pre-
0
trial site audits and research
−1000
1 48 95 142 189 236 283 330 377 424 471 evaluations of the crude feed
Observation – would eliminate the contin-
ual oil under-carry problem.
Figure 2a Improvement in first stage effluent oil and grease: consistent perfor- With the main operational con-
mance with Nalco Champion programme straint removed, the aim shifts
to optimising the desalting pro-
Base Incumbent NC Incumbent Nalco cess by following established
case trial trial trial II Champion
Nalco Champion M-O-C best
Filterable Solids

100
practice protocols. From there,
Removal, %

80
60
agreed-upon trial protocols are
40
followed to restore the expected
overall unit performance. At
20
1 48 95 142 189 236 283 330 377 424 471 that point, the programme’s
Observation robustness can be fully tested
with increasing amounts of the
Figure 2b Improved solids removal efficiency site’s main heavy Canadian
opportunity crude. Table 1
Base Incumbent NC Incumbent Nalco shows an overview of the unit.
case trial trial trial II Champion
100
efficiency, %

Solution
Desalting

80
Nalco Champion first con-
60
ducted thorough site audits to
40
gain insight into current oper-
20
1 48 95 142 189 236 283 330 377 424 471 ating practices, review base-
Observation line performance, and evaluate
tank farm and desalting equip-
Figure 2c Increased desalting efficiency ment to try to understand
whether there were any design
Canadian crudes that offer fur- formance had deteriorated flaws or mechanical limitations
ther discounts. markedly as the refinery intro- that could be responsible for
duced new and challenging poor performance. These audits
Case study 1 feedstocks to improve operat- also included understanding
A North American refiner had ing margins. Significant capital impacts to wastewater treat-
been working with an incum- expenditures had been made ment operations.
bent process treatment provider to improve perceived equip- Audit findings were then
for many years. Over time, per- ment shortfalls at the waste coupled to the research evalua-

2 PTQ Q3 2017 www.digitalrefining.com/article/1001414

tions and the company’s experi-
ence of successfully processing
Base Incumbent NC Incumbent Nalco
case trial trial trial II Champion
heavy, high solids-laden crude

Charge rate,
140000
slates. Our recommendations

BPD
120000
then established an effective
desalter monitoring and sam- 100000

pling trial protocol to allow for 80000

continuous evaluation of the 1 48 95 142 189 236 283 330 377 424 471
Observation
correct areas. Along with this,
best practice injection locations Figure 2d Higher average crude charge (while maintaining KPIs)
and safe staging areas for prod-
uct storage in the unit were also NC Incumbent Nalco
identified.

Advantaged crude,
trial trial II Champion
60
The Resolv treatment pro-

% in blend
45
gramme chosen utilised a cur- 30
rent best practice primary 15
emulsion breaker (EC2472A) in 0
conjunction with an enhanced 1 32 63 94 125 156 187 216 249 280 311
solids removal aid (EC2600A). Observation
Nalco Champion Research ini-
tially developed the EC2600A Figure 2e Significantly higher rates of advantaged crude in the slate (while
to meet the growing challenges maintaining KPIs)
of heavy, high solids crudes,
focused on Canadian dilbits. the opportunity crude level elevate refinery margins if the
Today, this combination pro- reached maximum rates above potential downsides can be mit-
gramme has an extensive treat- 40% due to hydraulic limits on igated to deliver consistent per-
ing library for reference cases the unit. Since Nalco Champion formance. The integrated team
like this. started to treat the unit perma- approach and systematic pro-
The primary goal from the nently, the amount of advan- tocols described above allowed
start was to achieve a new base- taged crude has averaged more for sound mitigation strategies
line performance (new sys- than 30%, while maintaining or to be developed and utilised
tem steady state) with no new exceeding past desalting and while reducing the operating
upsets. The first trial involved solids removal KPIs. risks. To make a step change
EC2472A being used with the Figures 2a-e highlight some of while maintaining high lev-
opportunity crude in the slate the performance gains in terms els of availability requires this
at the historical maximum of significant reductions to work and planning to be done
rate processed – only 2-3%. desalter effluent oil and grease in a consultative manner.
Oil under-carry immediately levels (COD in the total influ- There are chemical treatments
ceased (see Figure 2, ‘NC Trial’ ent to wastewater also dropped and monitoring tools and tech-
legend). The incumbent pro- as a result), increased solids niques that can be employed
vider was allowed a second removal and desalting efficien- to do this safely and hold the
test run, but the old perfor- cies, higher crude charge with gains when running these kinds
mance returned. As was men- more advantaged crude being of advantaged crudes.
tioned earlier, a key component processed. It is also important The desalters are now pro-
of these tests was to determine to note that these performance cessing higher levels of oppor-
robustness in the treatment pro- improvements are statistically tunity crude at new record
gramme. Consequently, dur- significant events, shown by charge rates. The average
ing a second test the amount P-values of less than 0.05. opportunity crude processed
of advantaged opportunity is now around 33% (a ten-fold
crude significantly increased Value increase compared to the base-
as part of the crude slate to test Bringing in higher levels of line) and the unit has hit a new
robustness. During this period, opportunity crude can greatly record run rate above 150 000

www.digitalrefining.com/article/1001414 PTQ Q3 2017 3

 Case study 2 system parameters
b/d without incident or upset.
Estimated gains in operat-
ing margins brought to the site
Design charge rate, bpd 250 000 [historical ~230 000]
Crude blend gravity, API 20-21° [historical ~24-27] have been calculated at $50-80
Design wash water rate, % of crude charge 8.0% million. Now previously ‘black-
Desalter design and configuration Petreco Bilectric; two-stage in series; two trains balled’ advantaged crudes can
Residence time, minutes Oil = 14-15; water = 90-95
be run with greater confidence
to achieve higher margins with
Table 2 reliability and KPIs intact.

I-MR chart of effluent O&G

Case study 2
Incumbent Nalco
Another Midwest US refin-
Champion ery was struggling to process
Individual value,

5000
its design crude slate which
2500 was 100% heavy Canadian.
ppm

Although the causes and effects

0
of poor performance were dif-
1 19 37 55 73 91 109 127 145 153 181
ferent to those of the first case
Observation study, the determining factor
was again lack of desalter treat-
I-MR chart of desalting efficiency ing robustness.
Incumbent Nalco The problems associated with
Champion
this unit were more compli-
Individual value,

100
cated than in the first case. The
95
crude unit was relatively new,
but mechanically the desalter
%

90

85
vessels and transformers capac-
ities were built correctly in
1 19 37 55 73 91 109 127 145 153 181
terms of meeting the planned
Observation
charge rates. However, this unit
was not suffering from obvi-
I-MR chart of desalted crude
ous problems like oil under-
Incumbent Nalco
Champion carry for example. Rather
Individual value,

1.00
there were layers of issues that
0.75 would reach a tipping point
wt%

0.50 on a semi-regular basis, caus-

0.25 ing emulsion growth, desalter
0 grid trips, significant water
1 19 37 55 73 91 109 127 145 153 181
carry-over, and finally frequent
Observation
crude heater trips, very costly
events. Additionally, regular
I-MR chart of dehydration efficiency fouling and cleaning of efflu-
Incumbent Nalco ent heat exchangers occurred,
Champion
Individual value,

100 and big problems arose at the

wastewater treatment plant.
95
The system parameters are
%

90 shown in Table 2.

1 19 37 55 73 91 109 127 145 153 181 Solution

Observation First, Nalco Champion con-
ducted a thorough unit audit
Figure 3 Desalter KPI improvements in case study 2 and followed a protocol sim-

4 PTQ Q3 2017 www.digitalrefining.com/article/1001414

ilar to that in the first case.
Laboratory screening proce-
dures again showed that the 7 Try line, 3 Jul 15
Try line, 3 Sep 22

Resolv treatment programme

of choice for these kinds of 6

Centrifuged solids, %
heavy Canadian crudes was 5
to combine a primary emul-
4
sion breaker (EC2472A) with Try line, 3 Oct 13
an enhanced solids removal aid 3
Try line, 3 Sep 15

(EC2600A).
2
There is a key point to under-
stand with this case. The refin- 1
ery technology group invested
0
a lot of time and energy to

15

15

15

15

15

15

15
01
determine if any mechanical

20

20

20

20

20

20

20
l2
n

ct

ct

ov
Ju
and operational deficiencies Ju

Au

Au

Se

N
15

23
25

13
4

24

12
existed that could be the cause
of historical poor performance,
before eventually advocating a Figure 4 Emulsion quality changes viewed through a contrast microscope,
desalter trial of a new treatment down to 10µ
programme. Key mechanical
issues that were fixed prior to KPIs. The performance gains the blue boxes represent incum-
the trial were: in three of these criteria were bent treatments on the unit.
1. Transformers: although sized substantial and represent a sta- • 15 July (incumbent): corre-
correctly (3 x 250 kVA per stage) tistically significant change lates to a wide emulsion with
with adequate secondary volt- (p-value = 0), whereas desalting a lot of solids ‘hung up’, as
age settings, they were found to efficiency had no loss in perfor- shown by the percentage of
be incorrectly wired. This was mance (P-value = 0.87): centrifuged solids on the y-axis.
corrected first. • Effluent oil and grease were You can clearly see the water
2. Level control: using a sophis- reduced by more than 50% droplets coated in oil and sol-
ticated nucleonic density meas- • Water in the desalted crude ids – unable to coalesce as they
urement, this was slow to react immediately dropped by 0.3% should do when under the
to density changes that span the • Dehydration efficiency influence of a DC field from the
desalter profile. Since the level jumped by 3-4%. desalter grids. Lack of emul-
control is linked to the brine An extra, higher level test sion resolution was deter-
valves, this was an important was also performed to see what mined as one of the key ‘bad
factor in reacting to emulsion was going on inside the emul- actors’, leading to the lower
growth. Larger sources were sion band. With the use of a grid becoming stressed and
installed to achieve the desired contrast microscope, the same tripping, then water carryover
refresh rate in the system. sample point (#3 try line) was would take place, and the crude
As with any trial, the first taken throughout the trial, heater would trip.
rule is always to transition to including periods when the • 15 September (Nalco
the new treatment programme incumbent was treating. The Champion): after the pro-
without causing any upsets. #3 try line was chosen as it was gramme acted on the unit for
Once this is achieved and a historically where the emulsion approximately two weeks,
new steady-state operation is was always present. the emulsion is completely
reached after the first few days Figure 4 shows a summary resolved. It is no longer present
of operation, the agreed-upon of some of the key samples at this location, leaving crude
trial protocols can start to be showing quality changes in the with only a handful of water
actioned. emulsion between the incum- drops.
Figure 3 shows the first six bent programme and Nalco • 22 September (incumbent):
months of data for the typical Champion’s. The areas inside a brief return to the old pro-

www.digitalrefining.com/article/1001414 PTQ Q3 2017 5

 gramme showed how quickly many tens of thousands of bar- level of performance improve-
the emulsion build is, with sol- rels above historical levels. ment is to be reached through
ids again ‘stuck’, helping to sta- The site estimated the annual automation, decisions and
bilise the emulsion. value attributed to switching to actions can be made because
• 13 October (Nalco Champion): a more robust desalter manage- system KPIs are in check and
shows again the emulsion being ment programme at greater than process variables mitigated.
fully resolved. $100 million. There is a saying by Lord
The point here is that the Kelvin: “You cannot control
treatment programme of Conclusion what you cannot measure”.
EC2472A and EC2600A is well Not all desalter treatment Part 2 of this series will take
suited to the challenges that chemistry is created equally this point a step further: you
heavy Canadian crude slates and theoretically, if the desalt- cannot change what you can
often present. By studying the ing equipment is properly measure without assurance of
emulsion down to the micron sized and specified for the stability.
level, it was clear that the sol- intended crude slate, should
ids were being successfully only be viewed as a polish- Brad Mason is a Marketing Director with
transferred to the brine, exactly ing step. However, these two Nalco Champion, overseeing the Process
where they need to go. cases demonstrate the substan- Global Marketing group for Downstream,
tial improvements a sound best managing the development of the
Value practice chemistry approach refining and chemical processing
This refinery’s crude unit did can produce on a crude unit portfolios. With over 21 years of
not have a single upset during struggling to meet performance experience, he holds a BTechnology
or since the trial and the pro- metrics and financial targets. degree in geochemistry from Macquarie
gramme has been installed on a Therefore, it is critical to have University, Sydney, Australia.
permanent basis across all of the a thorough understanding of
crude units on site. For the trial the unit so that a tried and true
unit, the desalters now regu- management programme can LINKS
larly process the designed 100% be applied and evaluated to
heavy crude slate, and often at deliver results. Utilising best More articles from the following
above name plate charge rates, practice chemistry can lead categories:
as high as 290 000 b/d. In sim- to stable operations, regard- Corrosion and Fouling Control
ple terms, the site can run the less of the crudes being pro- Crude and Vacuum Units
crudes they want/need to at cessed, such that when the next

6 PTQ Q3 2017 www.digitalrefining.com/article/1001414