One of the few refinery mercury management case studies available that I jammed out for HP (have to give some credit to some of the former founders of MMS - now PMG). This case study is truly one of the building blocks of my career and led to the improvement and development of several technologies that started when I was at PSC working with one of the giants and pioneers in mercury fate and transport in hydrocarbon process systems (Dr. Mark Wilhelm - RIP my friend).
Unblocking The Main Thread Solving ANRs and Frozen Frames
Case study: Refinery mercury chemical decontamination in preparation for decommissioning
1. Environment
and Safety
R. RADFORD, Portnoy Environmental Inc.,
Houston, Texas
Case study: Refinery mercury chemical
decontamination in preparation for decommissioning
Initial mercury management strategies began in 2010 at a US
refinery processing approximately 200 Mbpd of North Slope
crudeoil.Duringmaintenanceandinspectionactivitiesin2010,
elemental mercury was observed in hydrocarbon processing
vessels throughout the solvent extraction unit (SEU) (FIG. 1),
with interior vessel mercury vapor concentrations greater than
500 micrograms per cubic meter (µg/m3
).
Based on this observation, studies were conducted to assess
the distribution and accumulation of mercury in primary
process units that were previously scheduled for maintenance
and inspection operations. Mercury mass loading rates in
steel process systems were integrated with data from the
measurement and monitoring of mercury in process streams
and wastes to develop a robust mercury flux model. Effective
mercury management and chemical decontamination
plans were developed based on the model and successfully
implemented in 2011.
Subsequent mercury mass flux, distribution and chemical
reduction studies were conducted in 2013 and 2014 to
support chemical decontamination in preparation for refinery
decommissioning. Bench-scale decontamination tests were
conducted with process pipe coupons from crude distillation
unit 1 (CDU 1; operational for 20-plus years) to evaluate the
efficacy of mercury-removal technologies.
Mercury mass loading rates, speciation and depth profiles
from test-section pipe coupons were integrated with data from
themeasurementandmonitoringofmercuryinprocessstreams
to develop an accurate mercury flux model. Understanding the
nature and distribution of mercury, along with depth profiles
in carbon and stainless steel (SS) process equipment, is
critical to developing effective chemical decontamination and
decommissioning plans. Furthermore, understanding mercury
accumulationratesandthedistributionofmercuryandmercury
compounds is required to select the correct decontamination
and processing technologies. Shutdown plans were developed
based on the new model and implemented with complete
success during May–June 2014.
Thiscasestudypresentsasummaryofmercury-management
strategies and technologies required for mercury chemical
decontamination of impacted hydrocarbon processing systems
in preparation for decommissioning.
Applications. Throughout the oil and gas industry, the impact
of mercury in produced hydrocarbons is becoming more of
an emergent issue. This is not only the case for production
from certain unconventional resource plays, but also for assets
processingconventionalproductionastheyneartheendoftheir
economic life and as process systems require decommissioning.
Produced mercury contaminates multiple hydrocarbon
processingsystems(i.e.,upstreamproductionassets,midstream
gathering and fractionation plants, and downstream processing
plants) for which the dismantling, removal and disposal
presents unique challenges and risks to decommissioning
personnel and to ecosystems. Global conventions provide a
framework and guidance for decommissioning of oil and gas
facilities; however, specific regulatory guidance on residual
mercury concentrations that can remain in production systems
(scale or complexed in steel) is not available.
As the recycled metals value of hydrocarbon processing
plants can be considerable, it makes sense to remove
mercury to an acceptable mass/surface area that will allow
for safe transportation and recycling. The planning of safe,
environmentally responsible and cost-effective maintenance
Naphtha feed
STAB 2 BTMs
Naphtha 2 raffinate
Lean solvent
Rich
solvent
Naphtha 2
extract stripper
Naphtha 2
extractor
892.14 ppb
Naphtha feed
STAB 1 BTMs
5.47 ppb
V-04303
V-04305
V-04308
205.87 ppb
OVHD recycle
0.73 ppb
FIG. 1. Elemental mercury was observed in hydrocarbon processing
vessels throughout the solvent extraction unit.
Originally appeared in:
November 2015, pgs 61-66.
Used with permission.
HYDROCARBON PROCESSING NOVEMBER 2015
2. Environment and Safety
and decommissioning of mercury-impacted oil and gas facilities
isimprovedwithanaccurateassessmentofmercurydistribution
and the evaluation of applicable mercury-removal technologies.
Results, observations and conclusions. Studies indicate
that the mercury mass-loading potential of steel pipe exceeds
estimates reported in previously published studies. Thermal
desorption and chemical reduction bench testing indicate that
theprocessisreversible,usingvariouschemistriesandmethods.
Also, recent thermal desorption experiments performed on
metallic test coupons in a quartz tube furnace indicate that
field steam-out temperatures (100°C to 200°C) are ineffective
in removing mercury from steel, but may still be effective in
volatilizing hydrocarbon-soluble mercury and volatile mercury
species.
However, these species represent the smallest fraction
of the total mass of mercury within the scale and steel. This
can be precisely measured to quantify mercury emissions to
atmosphere from the flare during degassing and chemical
cleaning. Mercury removed from hydrocarbon processing
systemsduringsteamingandchemicalcleaningcanbeadsorbed
on media and removed from liquids so that the spent chemistry
is rendered nonhazardous and suitable for routine disposal.
Mercury vapors desorbed during steaming of hydrocarbon
processpipingandvesselscanalsobeadsorbed/chemisorbedto
various media, preventing release to the environment. Reactive
fluid beds and reactive sorbent media used in both cases remain
a waste streams that must be managed.
Technical contributions.Datafrommercurymassfluxstudies
using modified and improved methods, and integrated with the
results of metallic coupon mercury mass loading, distribution
and chemical reduction testing, has led to the development of
advanced chemical decontamination methods and chemistries
that are effective in removing 99% mercury mass and oxide
scale (depth profile of 1 mm), deactivation of pyrophoric iron
and encapsulation/removal of hydrocarbons.
With the development of new mercury-removal chemistries
came the development of new analytical methods to measure
the performance of these chemistries during chemical
decontamination and processing (waste minimization).
Laboratory and field trials of the analytical method led to an
improved understanding of economical and efficient mercury
waste-minimization processes effective in removing mercury,
metals, hydrocarbons and other contaminates.
Mercury mass loading, distribution and chemical reduction
bench scale studies are a key component of evaluating
decontamination chemistries/methods and identifying the
most cost-effective technology for application to mercury-
impacted hydrocarbon processing systems. In an effort to better
understandandquantifytheadsorption/desorptionofmercury
in steel and processing, technical modifications to a proprietary
sampling system were designed to allow for connection to plant
flare systems to quantify mercury emissions from steaming and
chemical cleaning operations.
Crude oil laboratory comparison. The appearance of
mercury at processing facilities can be delayed by months or
yearsduetothescavengingofmercurybysteelpipelinesurfaces.
It should be noted that this refinery is located approximately
500 mi from the supply, and that the first turnaround where
condensed elemental mercury and volatile mercury were
recorded in process systems occurred in 2010. A previous
turnaround performed in 2005 on the SEU made no record of
observed mercury.
Although no mention was made of mercury in the 2005
turnaround, process stream sampling and analysis were not
performed to verify the presence of mercury in process streams
or equipment; however, mercury sampling was not standard at
therefineryatthattime.Alaboratorycomparisonwasperformed
on crude oil composite samples collected from the inlet feed to
gain visibility into mercury concentrations and trends in crude
oil processed by the refinery. There was some slight variability
in the data reported from the three laboratories sourced for the
study (4 ppb to 6 ppb); however, those concentrations are in
the range of a previous study performed in 2004.
Conclusions could not be made to support the duration of
pipeline equilibrium or the possibility that new production
could have caused a substantial increase in crude oil mercury
concentrations. However, mercury accumulates and
concentrates in process systems, and data from recent mass
flux studies support the conclusion that 4 ppb to 6 ppb of
mercury in inlet feeds is sufficient to produce process stream
concentrations greater than 1,000 ppb.
Mercurydistribution and chemical reduction.Turnaround
and decommissioning planning should consist of attempts to
determinetheextentandtypeofmercurycontaminationpresent
in process systems. Routine mercury assessment sampling and
analysis, along with mass flux studies performed throughout the
lifecycle of an asset, provide valuable information required for
decontamination and decommissioning planning.
Data from mercury mass flux studies integrated with the
results of metallic coupon mercury distribution, depth profiles
and chemical reduction studies provide planners visibility
into the process system and the means to design/modify
select cleaning chemistries, chemical cleaning flow paths,
temperatures and residence times based on accumulation rates,
results of functional and molecular speciation and process
design restrictions.
Two separate mass flux studies were performed in January
2011 and August 2013 to understand the mercury sorption
and distribution dynamics associated with process systems
scheduledfordecontamination/decommissioning.Inthattime,
the mass flowrate balance of the SEU in 2014 (process feeds vs.
process outputs) indicated a potential mercury accumulation
rate of 0.0003 lb/hr, which was an order of magnitude more
accumulation than in 2011.
Also, the concentration of mercury in the naphtha feeds to
the SEU increased from approximately 5 ppb in 2011 to roughly
9 ppb in 2014, resulting in approximately 10 lb/yr of mercury
being introduced into the process. Both mass flux studies
clearly identified mercury sinks in the SEU, as well as net gains
in mercury in process fluids as they moved through the system.
As part of the mass loading study to support the 2014
chemical decontamination, a chemical reduction study was
performed to focus on the development of chemistry effective
in removing oxide scale, hydrocarbons and mercury from steel
HYDROCARBON PROCESSING NOVEMBER 2015
3. Environment and Safety
(FIG. 2). Once distribution, depth profiles, speciation and mass
loading per surface area had been established (test section
from C4 stream piping downstream of crude column overhead
receiver), select test coupons were subjected to 14 chemical
formulations and evaluated for mercury mass removal efficacy.
For each chemistry tested, two coupons (primary and
duplicate) were inserted into a treated SS chemical reaction
chamber and then subjected to each test case chemistry
for predetermined residence times. Process parameters
(temperature, dissolved iron, pH, residence time, total mercury,
flow, Re number) were collected every 20 min for the duration
of the chemical test. Precise measurements, weights and
microscope photos were taken of each coupon during inventory
into the laboratory and post-chemical testing. Several test-case
chemistries removed hydrocarbon, oxide scale and mercury
from the surface to 1 mm in 4 hr at a temperature of 50°C. This
is significant, as mercury mass loading rates were around 70 g/
m2
, which equated to around 35 mg of mercury per test coupon
that had to be affected by the chemistries.
The test case selected and deployed for the full-scale chemical
cleaning program in 2014 had an efficacy of up to 99%, coupon
mass loss of 2%, and depth of scale penetration of 0.51 mm.
Chemistries used in this study included modifications to existing
proprietary formulations, as well as some new formulations.
Chemical decontamination. Deposition of mercury and
its compounds can occur by adsorption, chemisorption,
precipitation and/or condensation. The amounts of elemental
mercury and its compounds in processed fluids affect all of
these mechanisms. The uptake of mercury in steel is primarily
through adsorption and chemisorption into the scale, making
both carbon and SS excellent mercury scavengers. With some
effort, this process can be reversed, depending on many factors
(what goes in can come out).
However, mercury complexed and incorporated into steel
surfaces is not easily affected by typical hydrocarbon chemical
decontamination chemistries and methods. Since 2005,
Portnoy Environmental’s research group has concentrated
efforts on understanding sorption dynamics of mercury
in steel pipe and in the development of effective chemical
decontamination solutions.
On a scale, some chemistries are more effective than others
for decontaminating mercury from hydrocarbon process
systems, and each requires careful consideration before use.
Generally, strong oxidizers and acids are the most effective,
but they come with corrosion risks and can require additional
processing steps to remove mercury from spent fluids. Mercury
can be oxidized by oxidants including halogens, hydrogen
peroxide, nitric acid and concentrated sulfuric acid. Using any
of these options for mercury chemical decontamination in high
concentrations, coupled with heat, should only be attempted by
qualified chemical and industrial services companies.
If the objective is strictly related to health and safety (e.g.,
for inspection), then the target may be to convert mercury to
a nonvolatile species rather than to remove it. If the goal is to
meet disposal requirements, then the decontamination process
may be designed to convert mercury to a nonsoluble species so
as to meet leachate criteria.
Total mercury removal is possible for systems scheduled
for abandonment and decommissioning, but as the chemicals
used for this purpose are aggressive to mercury, they typically
are to other metals as well. Less aggressive chemistries (aqueous
noncorrosive surfactant/chelant blends) are easy to process,
but, depending on mercury mass-loading rates and mass-
removal objectives, they can require higher temperatures,
increasedresidencetimesandmultipleapplicationtechnologies.
Reactive chemistries react with mercury to form water-soluble
or insoluble mercury (HgS), or otherwise combine with metal
ions. Since mercury is soluble in hydrocarbons to a certain
Flare
Mercury removal
vapor system
Mercury removal
liquids system
Naphtha 2 extract
stripper Hg compliance chemical liquid effluent
spill point
Compliance Hg vapor/liquid spill point
Compliance Hg hydrocarbon vapor
spill point
Solvent
recovery tower
Steam/chemical
injection
Steam/chemical
injection
FIG. 3. During mercury decontamination in both 2011 and 2014,
vapors and fluids were processed to minimize mercury emissions to
atmosphere and to remove mercury and other contaminates from
spent chemistries and condensates.
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5
Chemical test formulation number
Efficacy of test
formulations
Hgremoval,%
6 11 12 13 14
99.84
45.61
39.88
41.51
57.53
41.97
91.88
99.71
99.66
51.57
FIG. 2. A chemical reduction study was performed to focus on
the development of chemistry effective in removing oxide scale,
hydrocarbons and mercury from steel.
HYDROCARBON PROCESSING NOVEMBER 2015
4. Environment and Safety
extent (around 2 ppm), various surfactants and solvents are also
effective in dissolving mercury and in removing organic solids.
Generally, total mercury removal from a hydrocarbon
processing system is not considered unless the system is
scheduled for abandonment and decommissioning, or unless
process risk reduction is required. For example, the turnaround
in 2011 required systems with mercury accumulation rates
ranging from 1 lb/yr to 100 lb/yr to be decontaminated to
allow for extended maintenance activity over a 10-day period.
The chemistry selected for this purpose was noncorrosive
and designed to remove hydrocarbon-soluble and particulate
mercury and to form soluble and insoluble mercury salts.
The selected formulations of surfactants and chelants were
applied in vapor phase and cascade phase over 12 hr to 24 hr
with success in reducing interior mercury concentrations from
around 1,000 µg/m3
to less than 1 µg/m3
, as continuously
monitored from systems over the duration of maintenance.
The goal for the 2014 SEU decontamination was
decontamination in preparation for decommissioning,
which generally requires 99% mercury mass removal to
allow components to be processed at a metals recycler or
reused at another processing facility. Chemistry selected for
this purpose was a blend of surfactants (encapsulate metal
ions and hydrocarbons), mineral acids, corrosion inhibitors,
penetrants and chelants with a hydrogen sulfide scavenger.
In both cases (mercury decontamination for reuse
and decontamination for decommissioning), continuous
performance sampling and analysis were performed to provide
guidance on the effectiveness and duration of each chemical
phase. Assimilation of this data provides an approximate total
mass of mercury removed from each system. Because of the
difficulty in capturing all the fractions, a more accurate method
of determining mercury mass loss is by thermal desorption or
aciddigestionofselectsteelfromprocessandcomparisontothe
average mercury mass loading per surface area of pre-chemical
application. Another verification method being pioneered is
X-ray fluorescence analysis, which is yielding promising results
with correct software and custom calibration standards.
Spent chemistry and vapors processing. Chemical
and processing procedures can be applied to spent fluids and
condensates such that the materials are rendered nonhazardous
and thus suited for nonhazardous disposal. Medias used for
encapsulation, reaction or chemisorption, and filters used for
colloidal particulate removal during processing, still need to
be characterized and managed based on the results of waste
characterization.
Some commercial mercury-removal systems are targeted at
vapor-phase treatment, and some are targeted at liquids. Vapor-
phasetreatmentsystemsprimarilyconsistofsulfur-impregnated
carbon, metal sulfide on carbon or alumina, and regenerative
molecular sieve (zeolite), onto which is bonded a metal that
amalgamates with mercury. Liquid-removal processes consist
of iodide-impregnated carbon, metal sulfide on carbon or
alumina, and silver on zeolite molecular sieves.
Functional and molecular mercury speciations are not only
important to developing effective chemical decontamination
plans, but they are also important to planners for designing
effective processing systems.
In both cases (mercury decontamination for reuse in 2011
and decontamination for decommissioning in 2014), vapors
and fluids were processed to minimize mercury emissions to
atmosphere and to remove mercury and other contaminates
from spent chemistries and condensates (FIG. 3). An innovative
active fluid bed was designed and installed prior to solid media
beds to minimize media changeouts. This new approach has
exceptional mass-loading potential and cost benefits, and, when
coupled with a solid media system, performs extremely well.
As with chemical decontamination, continuous performance
sampling and analysis are required to provide guidance on the
effectiveness and duration of each phase of processing. As a
means of determining how accurate the field analysis methods
can be, an instrument detection limit study was performed with
the field analyzer on each chemical formulation selected for the
2011and2014events.Inaddition,fieldduplicatesandlaboratory
triplicates were analyzed for compliance of processing.
A total of 100 Mgal of spent chemistry/condensates were
processed (average pre-processing concentration of 500 ppb
of mercury) during the 2011 turnaround to less than 7 ppb
of mercury. As the chemistry used during the 2014 shutdown
was more complex (thereby removing significantly more iron
and mercury), additional processing steps were required to
completelyneutralizeandprocess400Mgalofspentchemistry/
condensates with average mercury concentrations ranging from
500 ppm to less than 4 ppb of mercury.
Compliance monitoring of flare. As a chemical element,
mercury is not destroyed by a thermal oxidizer, but is only
convertedtootherchemicalforms.Althoughsteamingequipment
may remove some volatile mercury species and hydrocarbon-
solublemercury,thetemperaturesaretoolowtoaffectcomplexed
mercuryinsteel.Manyhydrocarbonprocessingplantsarelocated
nearenvironmentallysensitiveareas,andmercuryemissionsfrom
steaming to flare and materials processing should be considered
and mitigated (FIG. 4).
A considerable amount of mercury can be released to the
atmosphere from steaming and chemical cleaning of mercury-
contaminated systems during turnarounds and shutdowns
FIG. 4. Many hydrocarbon processing plants are located near
environmentally sensitive areas, and mercury emissions from steaming
to flare and materials processing should be considered and mitigated.
HYDROCARBON PROCESSING NOVEMBER 2015