NPRA Q&A-Conclusion
This article is the last of a three-part series of excerpts from the 1996 NPRA Q&A transcript. It includes details on operational and mechanical techniques for desalting heavy crudes.
The first article in this series featured discussions of fluid catalytic cracking unit operations (OGJ, Oct. 13, 1997, p. 50). The second article included practical information on hydroprocessing, reforming, and light-oil processing catalysts (OGJ, Oct. 20, 1997, p. 58).
For more information on the panelists and the format of this meeting, see Part 1 of this series.
Heavy oil desalting
What specific techniques are being employed to desalt heavy crude oil (less than 20? API)? Review important process parameters (such as desalter efficiency, removal of filterable solids, etc.) and type of equipment in service.Lemmon: Tosco's Avon refinery effectively desalts heavy California crudes with an API gravity of 13.4°. Salt removal is excellent at roughly 95%, yielding a little less than 1 lb/1,000 bbl in the desalted oil. Basic sediment and water (bs&w) removal is also good, obtaining 85% removal, yielding less than 0.2%.
The water effluent, which has been a focal point of our refinery, is fairly clean at 0.5% oil. The equipment used is two generously sized Petreco low-velocity desalters operated in a series. The oil residence time is approximately 1 hr in each of the desalter vessels for this 60,000 b/d application. A brine-settling tank is also utilized to maximize oil recovery from the brine effluent.
The operating temperature of the crude is critical and must be maintained over 315° F. Wash water is added at 6-8 vol %, and the wash water source is vacuum tower ejector condensate combined with stripped foul water.
The pH of the wash water blend is about 5.6, and that is also critical with this high-naphthenic-acid crude. The pH must be below 6.5 to give acceptable performance.
We run the mix valves in the first stage at 4 psi, and at 7 psi in the second stage. One other thing we do is blend the crude with some low-solids recovered oil (less than 0.5% solids). This improves the API gravity to around 14.8° API.
We use emulsion breaker chemical at 20 ppm to 25 ppm, and the desalter runs very well. Over the years, the desalter operation has improved, and I attribute this to reduced solids in the recovered oil and today's more effective desalting chemicals. The last point is that we are careful not to blend crudes into the California crudes that could cause asphaltene precipitation.
Hunkus: In the 10-12° API range, several process techniques are employed to offset the difficulty of desalting heavy crude oils. These involve modified process parameters, special crude handling, and basic design considerations to improve results.
Elevated temperatures are very important to reduce the viscosity of heavy, highly viscous oils. Generally, an upper limit of approximately 325° F. is set due to the reliability of desalter electrical components. Sulfuric acid addition is generally used to control effluent water pH, which is usually held in the 5.5-6.5 range. HCl may have advantages, but it is more hazardous to handle.
You may need to consider elevating the gravity of the raw crude oil with recycle of atmospheric naphtha, distillate or other streams to increase the density difference between the oil and the water and to further reduce its viscosity. Low mixing energies are employed to counterbalance the high viscosity and low gravity of these crude oils so you can still achieve separation. Keep a log so the operators know where to "set the knobs" when this type of crude hits.
Try to maximize settling time in your crude storage tanks, and drain water as much as possible before charging to the crude unit (or coker, visbreaker, or heavy oil process). Check crude compatibilities in the laboratory and blend crudes by types and ratios that preclude asphaltene deposition. Keep records of blend recipes that work well in your plant for future reference. Accomplish tank switches over a period of several hours to preclude desalter upsets and minimize the potential water carryover on crude units.
For very heavy crudes, larger vessels and higher-kva-rated transformers are used to assist in overcoming the higher viscosity and lower gravities of these opportunity feedstocks. Consideration should be made for possible inclusion of special polymers to augment high-viscosity, low-gravity separation and conductivity. Several demulsifier systems can be included to allow switching to various systems without the need for recipe chemicals.
You should also allow for large rag layers and a good, well lighted sampling area. Finally, you generally do not want to run the heavy crude and water through a booster pump in the preheat exchange train. You would be better off buying a high-head pump and injecting the crude and water closer together right before the desalter.
Jackson: A number of our refineries in Australasia have residue cracking units (RCUs) and run a very wide range of crudes-Middle East sour crudes to Southeast Asian waxy crudes. The straight-run crudes have a feed specific gravity of between 0.83 and 0.90°, an API gravity of 39 to 26°.
These sites also run some imported residues through the units. As the imports tend to be wet, they are run through the crude distillation unit (CDU) desalters. With these regimens, they can run between 0% and 70% condensate with the residue. They have run specific gravities in excess of 0.93 (APIs less than 20°).
The combination of high portions of imported residues and low portions of condensates is the worst of all worlds, as the condensates precipitate out asphaltenes, which act as surface-active components and cause problems at the interface in the desalter. Some refiners are desalting atmospheric residue alone.
BP refineries with RCUs use multiple-stage desalting and no caustic injection into the hot residue as best practice. Current best practice is treating 14-16° API crudes with around 85-90% salt removal for a single stage desalter and 97-99% for a two-stage desalter.
The important parameters that need to be considered when running heavy crude oil are:
• Viscosity-The heavier crudes and residues are generally more viscous (although some of the lighter, waxier crudes can also be a problem from the viscosity point of view), and this becomes a problem with atomizing them, resulting in poorer mixing. It also adversely affects the coalescing of the droplets. If the viscosity can be reduced, then atomizing and mixing will improve and the residence time necessary to separate the fluids will be reduced. Viscosity reduction can be accomplished by increasing the temperature of the feed to the desalter and the desalter operating temperature, normally up to the limit of the bushings. This is typically a maximum of 300° F., although some new bushing materials are said to allow you to operate as high as 320-350° F. at 450 psig. The mixing can also be improved by increasing the pressure drop across the mixing valve. The increased mixing will also mean that the emulsion will be harder to separate, so desalter design and cooperation need to account for this.
• Density-The heavier crudes obviously have higher densities and are closer to the density of water. This makes the crudes that much harder to separate from the water. Consequently, more time and/or dewatering chemicals are required to separate the oil and water phases. The whole exercise is compounded by the fact that the heavier crudes do not dewater as well in tankage. So, the longer and better the dewatering in the tanks, the more efficiently they can be desalted. The result of water carried forward is more-stable emulsions, which will require specific and more demulsifying chemicals. The additional chemicals should be added before the crude hits the desalter. We typically predose some 2-3 hr before the feed tank change. The demulsifying chemicals basically neutralize the surface-active components that sit at the surface interface.
• Asphaltenes-These materials in the heavier crudes tend to be surface active and prevent the oil-water phases from separating efficiently. Again, surface-active chemicals will help the desalting process.
• Solids-The removal of solids while running heavier feedstocks is also more difficult. The solids will not move as easily from the oil phase into the water phase. You have to dose with suitable and larger quantities of the water-side polymers. These sit at the oil-water interface and enable the solids to pass more easily into the water. The polymers effectively wet the particulates. The polymers also have the advantage of aiding the deoiling of the water phase.
• Preheat-The preheating of the cold crude upstream of the desalters has to be considered. Most CDUs have good heat integration that preheats the cold crude pumparounds, overheads, product rundowns and sometimes cold residue from the CDU. The heavier crudes run a higher percentage of 680+° F. (360+° C.) material, so the temperature of the material entering the desalter can drop, resulting in poor desalting. This needs to be considered and designed for in the preheat system, while also allowing for fouling of the preheat train. Watch the velocities in the heat exchangers and make sure they do not drop too low.
• Wash water rate and quality-The wash water rate should be increased to improve the contacting of the oil and water streams. This should be increased up to a usual maximum of 7% on crude. The quality of the wash water is important. Wash water with a high pH can cause stable emulsions. This is normally an indication of caustic in the system, which can cause the presence of sodium naphthenates (soaps). The "soaps" can also be formed with naphthenic-acid-containing crudes with acid numbers greater than 1.0.
Most chemical vendors are capable of running bench-scale desalter simulators. These give an indication of those crudes that might be susceptible to poorer desalting; hence, the refinery and the chemical vendor can prepare an operating protocol for running these troublesome crudes and select efficient demulsifiers.
As can be seen above, the link the refinery has with its chemical supplier and the service and support that supplier gives are very important. A good service contract can make all the difference when running a wide ranging crude slate. The data base is an invaluable tool to troubleshooting problems retrospectively and then determining how to prevent those problems from recurring. The analysis of these data will also indicate whether you have a real problem with the desalter or an operational problem, such as not dewatering the crude correctly, generic crude-type change, spiking of other oils into the crude feed, etc. We have found that the historic data have helped us determine how much predosing of which chemicals we have to do for specific crude types and for any crude changes.
An interesting twist is when you have been running a crude for years and dosing successfully, and a new crude parcel desalts less efficiently. The chemical addition rates are the same, the desalter is checked out and appears to be functioning as expected, and the crude looks the same from its physical properties. Everything seems as it used to be, but this material is hard to desalt. We have found that some of the drilling-mud stabilizers used by the upstream exploration companies carry through in the crude and have been known to affect the desalting process.
In summary, the following techniques are used when processing heavy crudes:
- Dewater the crude in tankage as long and as well as possible. Give it time and attention.
- Predose the oil and water with chemicals well (2-3 hr) before the crude hits the desalter.
- Increase the operating temperature of the desalter up to the temperature limit.
- Increase the pressure drop across the mixing valve to try and increase atomization of the feed and water and increase mixing.
- Increase the wash water rate to aid contacting.
- Add wash water into the suction of the feed pumps upstream of the desalter to aid mixing, as has been done in some locations.
- Watch the oil-water interface with heavy oils as rag layers are easily set up. These will have to be blown down through the tricock system.
It is crucial that you understand the source of your raw crude. By that I mean getting back to what Mr. Jackson mentioned: Are they using downhole surfactants to stimulate the wells? We found out quite often they are, and that really can upset a desalter.
You also have to realize that on a typical heavy crude, the bs&w shipment specification is somewhere around 3%. Normally in lighter crudes it is 1%. That creates a lot more solids that you have to dispose of somehow. This leads to the next question which is based on an economic decision: Which way do you want to take those solids? Do you want to take them out with the water and dispose of them somehow or recycle them gently back in with the raw crude? (We usually do not advise that for heavy crudes.) Or, do you choose to leave them in the crude oil, assuming you have high enough velocities through your preheat exchange, and then have them end up in the coker should you have that particular unit? That is all based on economics for each individual refinery.
As for temperature, hotter is usually better in terms of desalting performance, but I caution everyone that this is not always the case. If you look at the Stokes Law equation, the difference in the densities between the oil and the water at a given temperature is what is really important.
Water densities will change fairly linearly, but not all crude oils do. Some have quite a curve to them. And what you are looking for is the temperature where that difference is the greatest. We have had a number of cases where we have actually had to lower the desalter temperature from where they were running, at 300-310° F. down to 280° F. to achieve the best desalting.
On the heavy oils, the other concern as you get hotter is that the conductivity of the crude oil increases. And, depending on the porphyrin content (the specific part of the asphaltene) you can very easily get hot enough to continually trip out your electrical grids.
The demulsifier selection is critical. In testing, you have to use the particular crude blend that is being processed at the time, as well as the wash water that is available at that particular desalter. I would also caution everyone that when you are looking at using solids-wetting technology, specifically polymer-type chemistries, not all polymers will drop enough solids. If that is what you are trying to accomplish with the particular polymers, you have to be careful as to what is selected.
Overall, you have to have a very good understanding of all the intricacies of what is going to affect your desalting performance. That goes all the way back to tank handling. And more critically, we are finding that if you get to the heavier crudes, what you do with the solid treatment is critical, as is what you do with the small amount of oil undercarry that seems to be the norm with the heavier crudes.
Barker: Our experience was that the pressure drop across the mixing valve should be monitored and controlled closely. Too high a pressure makes an emulsion that is hard to break, particularly if there are any solids present.
It is important to only use enough pressure drop to get good mixing. Increasing the temperature lowers the viscosity of the oil, and allows the pressure drop across the mixing valve to be lowered as well. However, that draws more current across the grid, and requires adjustments to keep from over-amping.
To overcome that, we installed a naphtha recycle (about 1,000 b/d with the 7,500 b/d of shale oil production) to lower the viscosity of the mix flowing through the mixing valve. All of this improved the desalter operation to the point that we had some control over the process.
J. L. Raina (Indiana Oil Corp. Ltd.): Mr. Jackson mentioned that water quality is very important. Could you throw light on whether it should be exclusively fresh water or can you recycle some of the process effluent from vessels? In two-stage desalting, can a little bit be partially recycled?
Jackson: Taking the last point first, yes we do recycle water from the second stage into the first-stage desalter at some locations. But remember, we are not processing the heaviest crudes.
To answer your question, contaminants like ammonia in the wash water can cause you real problems. So if you are carrying ammonia in your desalter, do not use these recycle streams when you are running these heavy crudes.
Also slops, sour water, caustic, or anything like that put into your crudes at the same time that you are into these heavy opportunity crudes are going to cause you real problems. Take these streams out, or at least minimize their processing, when running difficult crudes.
Lemmon: As I mentioned before, in our heavy oil desalter we are actually reusing condensate from our vacuum tower ejector systems blended with stripped foul water. The stripped foul water does have a little bit of caustic in it that we add to help strip ammonia. The trace amount of caustic does not seem to upset the desalter, and it is good clean water, low in solids.
J. L. Raina (Indian Oil Corp. Ltd.): Are there any special requirements for high naphthenic crudes?
Lemmon: The pH must be on the acidic side or you will form naphthionic soaps which will completely upset the desalter.
William Heimbaugh (Howe-Baker Engineers Inc.): One of the things we have found in our experience is that it is preferable to inject the crude-water emulsion just above the water interface level. This practice provides two explicit benefits when processing heavy crudes. First, it maximizes crude oil residence time. Second, it reserves the high-intensity electrical fields for the more difficult-to-resolve emulsions.
As Mr. Fearnside indicated, Stoke's Law dictates the water droplet settling rate inside the desalter. Typically, most of the water droplets fall out in the lower-intensity electrical field (in a Howe-Baker desalter) established between the bottom electrode grid and the water interface level. However, when processing heavier crudes, the water droplet settling rate is significantly reduced. Therefore, in order to remove the highly emulsified water droplets in heavy crudes, it is imperative that you reserve the high-intensity electrical fields (between the electrode grids) for this purpose only. This design practice ensures that the water content in the high-intensity electrical fields is minimized.
With regard to desalter operating temperature, there are a lot of people who think "the hotter the better." However, there are some limitations. Although some of these have already been mentioned, two deserve mention here.
First, as the operating temperature increases, the crude conductivity increases. The additional conductance of the feedstock places increased demand on the transactors and, accordingly, power consumption increases. If the power consumption becomes excessive, the transactors may trip off-line due to an overload condition.
Second, there are some equipment limitations with regard to temperature. In most desalter designs, the electrode insulators and entrance bushings are made from Teflon. These components typically have a withstand temperature rating of 375° F. (190° C.). Accordingly, we do not recommend prolonged operation above about 305-310° F. on heavy crude slates.
Robert Brierley (Husky Oil Operations Ltd.):
I have been working 15 years on desalters with Lloyd minster crude (21° API). The crude comes in with asphaltenes already destabilized, and we definitely have asphaltenes in our interface. Our interface pad under normal circumstances would grow and come out the bottom. We found the polymer addition program was a benefit to controlling that interface pad.
You have to be careful about the polymer you use. They can actually cause more asphaltenes to come out the bottom. Our asphaltenes-to-solids (clays) ratio has been as high as 4-to-1 in the desand sludge. Whereas if you look at it under normal circumstances in the pad, it can be 1-to-1.
One critical thing is water-quality control. We have actually taken our overhead water out of the desalter wash water system because it carried with it clays that were actually stabilizing the emulsion in our first-stage desalter.
I found working on parallel desalters at a refinery-Husky's other plant-that we have 97% salt removal with single-stage desalting. We went to two-stage desalting at our BPU unit; we saw 97% salt removal, no improvement. The reason was that we had to back off on our mix valves because of the increased tendency to help stabilize emulsions in the interface.
What works best for us so far is fresh stripped sour water addition to both desalters. If you can afford the higher water rate, put parallel wash water into number one and number two desalters at the same time, and your interface pad will come under control much faster with lower chemical dosages. We have been averaging 0.11% oil undercarry with single-stage desalting and somewhere around 1% with two-stage desalting. These are two different refineries and two different units.
Tom Collins (Petrolite Corp.): I agree with everything that has been said. Just one other thing to add is that the sediment removal is very important with desalting heavy crudes. Also, the techniques employed tend to vary with the design of the desalter.
The Petrolite Bilectric design allows for more intense mixing, which is a critical factor in desalting heavy crudes. Petrolite has extensive commercial operating data and pilot plant research that shows Bilectric design coupled with a proper chemical program produces much greater removal of salt and filterable solids.
Mechanical alterations
What kinds of mechanical desalter internal modifications are refiners making to better handle these heavy crudes? What is the state-of-the-art design for heavy crude desalting?
Jackson: As indicated in the first question, we have some experience with desalting heavy and residual feedstocks in our Australasian refineries. Most locations have modern, two-stage desalters, and one has a three-stage desalting system.
The answers to the previous question also indicate the problems, and some solutions, to heavy hydrocarbon desalting. From a mechanical point of view, the following are items that should be considered:
• For processing heavy hydrocarbons, a larger treating area per barrel processed is required. For example, a desalter designed for 60,000 b/sd of 36° API crude would be able to process 50,000 b/sd of 30° API crude, 36,000 b/sd of 26° API crude, and 14,000 b/sd of 18° API crude with the same efficiency; or maintain throughput at lower efficiency.
• The electrode bushing material should be looked at and replacement considered to allow the desalter to run hotter. Current best available is around 350° F. at 450 psig (177° C. at 31 barg).
• A fixed, in-line mixer can be installed, but we generally still use a mixer valve in series.
• Look at debottlenecking the hydraulics to the desalter on both the crude and the water side to enable higher pressure drop across the mixing valve and higher wash water rates, if these parameters are low.
• The heat exchanger network should also be reviewed if this is a bottleneck. The minimum velocity of the crude in the tubes should be noted and redesign of the exchangers considered, to prevent fouling of the preheat heat exchangers during heavy crude campaigns. Inserts can be considered.
• Each vendor has his own state-of-the-art design. Petrolite, for example, has its Bilectric design, which we have in some of our U.S. and Australasian refineries.
• A sludge wash-down facility should be included in the design to allow on-line removal of any sludge buildup from running heavy crudes.
• The water handling system has to be large enough to cope with any increase in water disposal due to increased wash water rates.
• Look at chemical addition facilities and talk with the suppliers to determine the rates and location for adding heavy crude desalting chemicals.
• Look at sample points and instruments to ensure you have sufficient data gathering facilities to enable a data base to be gathered and the future operation of the desalters on heavy feeds to be improved.
• Most vendors believe that closely spaced grids are important (6-12 in.), but the ability to change the grid spacing has reportedly been shown to not be an important process variable.
• All desalter vendors supply internals that vary from the simple single grid supplied from a single transformer to a triple-transformer system, each transformer supplying a separate grid. Again, different vendors have their own unique designs. Each design has its strengths and weaknesses. To date, BP does not have an absolute preference for desalter internal design.
• The inlet distributor design is claimed by some vendors to have an effect on the efficiency of the desalting.
• I would also look upstream and at other units feeding the desalters.
Other things I would consider to ease the burden on the desalters and make heavy crude desalting as easy as possible are:
• The crude/residue tank dewatering facilities and protocol. If water and particles do not get into the desalter in the first place, they cannot cause problems.
• The slops and caustic disposal strategies should be reviewed. Some sites, for example, that run a wide variety of crude types, do not run caustic or slops to their desalters when they run heavy crudes, as this only exacerbates the difficulty of the desalting of these difficult crudes. They have sufficient storage to process them when lighter, easier-to-desalt crudes are run. This seems to be successful.
• Look at the sour water stripper operation and check to see if high-pH water is being fed to the desalter.
Arndt: A big factor is the pH of the water in separating the oil from the solids. The right pH is crude-dependent.
Regarding the mechanical aspect, we favor the low grid velocities and bielectric grids for heavy oil desalting. The distributor can be a problem here because the distributors on Bielectric designs have very small gaps, about 1/4 in. in our application. If you go 20 years and never clean them, your desalter will not work very well.
Fearnside: One thing I would like to add is that, with the heavier crudes, you directionally have to handle a lot more solids. It is very important that your mud washers or mud-busting equipment be up to specification, capable of delivering enough water at the right velocities through the inlet header nozzle holes.
Gentry: Kellogg recently performed design, engineering, procurement, and construction for a desalter to process heavy crude of 11.6° API. It was necessary to dilute the heavy crude. A number of tests were done to determine the best diluent, and economic studies were done to support these tests. In the end, the heavy crude was blended with a lighter crude for a blended gravity of 14° API prior to desalting.
Pilot test runs were performed by the desalter vendor on bench-scale equipment to confirm that the following design desalting objectives will be achieved:
- bs&w, maximum 0.8 vol %
- Salts, maximum 15 ppm(wt) or 5 lb/1,000 bbl (80% salt removal)
- Sodium, maximum 3 ppm(wt).
The design desalting temperature for this crude blend was 300° F.; wash water injection rate was 10 vol %. Wash water will have less than 100 ppm ammonia, low hardness, and a pH of 6-8. Provisions for oil-soluble emulsifier injections were made at the suction of the charge pump. A water-soluble wetting agent is injected upstream of the mixing valve.
The vendor's double-ported, quick-opening type mixing valve was provided with the design. A capacitance-probe type interface-level transmitter was provided for interface-level control service. At the client's request, Agar Corp. probes were installed as backups.
This unit has recently been started up and operated on crude blends that are somewhat lighter than the design. Unfortunately, there has not been a good test of the heavy blend operation as yet. We would hope to have more feedback within the next 2-3 months.
William Heimbaugh (Howe-Baker Engineers Inc.) A couple of mechanical modifications come into play on these heavy crudes. One of them has already been mentioned, i.e., mud wash system for enhanced solids removal. I will caution you, if you have that kind of a system, you need to have a water effluent header to go along with it.
The second modification is what we call a cuff header for extraction of rag layers. One of the panelists mentioned rag layers being a problem. The header allows you to extract the cuff layer at that same elevation, rather than having your operators lower the water level, causing a temporary upset condition.
With regard to electrode design for heavy crude desalting, Howe-Baker recommends our state-of-the-art Enhanced Deep-Grid Electrode technology.
Copyright 1997 Oil & Gas Journal. All Rights Reserved.
