Search

HODGES (Athlon Solutions)

High iron content on desalted crude manifests itself in two main areas. These are inability to make anode-grade coke and as a FCC catalyst poison. The iron can be in many forms. Three of the most common forms of iron we see are iron oxide, which is rust; iron sulfide, which is corrosion; and, siderite, which is essentially crystalline iron carbonate that we are seeing more and more of in certain tight oils. Historically, operators and process chemical suppliers have tried to address iron removal in three ways. The first way we have seen is acidifying the washwater with citric, acidic, or glycolic acids. A second way is applying a chelant to the washwater such as EDTA (ethylenediaminetetraacetic acid). The third way is running a continuous dirty rag draw from the desalter. We have not seen any of these mentioned methods meet with much success with respect to iron removal, and they have the unpleasant side effect of increasing desalter, crude tower, and downstream unit corrosion. The handling problems of the continuous dirty rag have diminished the popularity of that method as well.

We have seen a very successful alternative approach where caustic is added to the desalter washwater along with a unique demulsifier, which functions well at all pH levels. This transports the iron into the water phase, thus exiting with the brine. Check with your current chemical desalter supplier before trying this approach.

JIM PROROK (Husky Oil Operations Ltd.)

For years, we have been successfully running the dirty bleed pull on the rag draw, and we continue to do so. Yes, we are an anode coke business, and good iron removal is essential. Maybe the difference is that they were a light sweet refinery. Nowadays, the solids contents are coming up in the crudes, but we still manage them. You are right; we must have it. The oily bleed has to go to a free water knockout drum and then get as much free water out as possible. The resulting oil stream has to be centrifuged.

HODGES (Athlon Solutions)

It can be managed; it is just a bit cumbersome.

SLOLEY (CH2M Hill)

What pH do you run the desalter water down when you add caustic for the iron partitioning in order to promote iron partitioning into the water?

HODGES (Athlon Solutions)

The pH of the effluent brine must be 9 or higher. You could actually do some studies of pH elevation versus your iron removal to determine the specific best target for your particular desalter.

ROBERT AJILUNI (Athlon Solutions)

I want to add that if the source of your washwater is from a sour water stripper, you will need make sure that the stripper is working well because sometimes that water will have an elevated pH due to ammonia. So, if you are checking your washwater or brine as it comes off the desalter, you would think: Oh, the pH is good, and I have plenty of caustic. But in fact, you do have not enough caustic; so, you are not actually removing the iron the way you should. It is such a fine line on the color of that brine. You do not want crystal-clear brine, but you do not want oil; so that it is really incumbent on the operator, and Operations in general, to walk that fine line. But if you walk that line, you can get some really amazing results. 

JILL BROWN BURNS (Valero Energy Corporation)

We actually do this at one of our facilities and have not really seen the benefits on most of our critical KPIs (key performance indicator). It has actually caused some difficulties. If you go back to one of our previous slides talking about controlling the crude overhead pH, you will see that this is the KPI we are not able to meet. Mr. Sloley indicated controlling the brine pH. It does go up when you add caustic to it, and then you see a subsequent issue with controlling the crude overhead pH below the upper specification.

GLENN SCATTERGOOD (Nalco Champion Energy Services)

Some contaminant removal additives (CRAs) can be added to the desalter washwater to make the iron more water-soluble (such as acids) and move it from the oil phase into the water, leaving the desalter, along with the brine water, to the wastewater treatment plant in a water-soluble form. Other CRAs are non-acidic and can be used to move particulate iron from emulsion layers in the desalter into the water phase.

CHRIS CLAESEN (Nalco Champion Energy Services)

If the largest part of the Fe is contained in the solids, the solids removal can be improved with a specific Nalco Champion solids wetting additive. Removal of over 90% of the Fe and desalted crude Fe levels below 0.5 ppm have both been achieved with this additive.

DENNIS HAYNES (Nalco Champion Energy Services)

Mitigation of iron carryover from the desalter depends on the form of the iron; but where it is iron sulfide particulate, increased washwater, optimized mix valve setting, addition of solids removal chemistries, and adjusted interface level are the most certain methods.

PHILIP THORNTHWAITE (Nalco Champion Energy Services)

The key consideration here is the form of the Fe. In the majority of cases, iron is present as iron sulfide, a corrosion byproduct which is present in the crude due to corrosion upstream, tankage, or introduced via external streams such as slops. The particle size of the FeS (iron sulfide) in the crude is very small; and when coated in oil, these particles are very difficult to remove at the desalter. In order to effectively removed inorganic iron present as particulates, specialist solids removal chemistries have to be used in conjunction with the primary demulsifier. When the oil and washwater pass through the mix valve, the adjunct products help remove the oil coating from the particulate, allowing it to be water wetted. Once water-wet, the solids are removed with the effluent brine.

Iron can also be in the form of iron naphthenate and is present in particular naphthenic acid crudes. This organic form of iron is effectively removed through acidification of the washwater. The reduced pH causes the iron to dissociate from the naphthenate molecule, leaving the naphthenic acid and the iron salt of the particular acid that was used. This iron salt is typically water-soluble and removed with the effluent brine.

PRIBNOW (CITGO Petroleum Corporation)

It is essential to keep crude pre-heat exchangers clean. Cleaning benefits are hard to justify if they negatively affect crude rate. The criteria we use are crude hydraulic throughput, heater firing limits, fuel gas savings, desalter temperature efficiency, or crude or vacuum tower heat balance if you have needs in that area to remove heat. Each of these criteria at CITGO is periodically monitored by the process engineer or unit engineer and unit personnel to determine if action needs to get taken. If the fouling factors increase, we will increase monitoring of that exchanger and begin to make plans to take it offline in the near future or next opportunity.

The most common methods we use at CITGO are rather simple: pressure surveys and U-factor or duty calculations. For the crude hydraulic restrictions in your pre-heat, perform pressure surveys. Find out what bundle has the greatest impact, or which will give the crudest throughput recovery, and then clean it. The exchangers should be ranked in order of greatest to least impact. Compare the design conditions with the start-of-run conditions. If bypasses around exchangers are utilized to increase crude throughput, downstream relief capacities must be checked, and heater inlet temperatures monitored.

Perform U-factor or duty calculations for heater firing limits, desalter temperature, and fuel gas savings. Rank the exchangers by the most fuel gas saved at the heater. If you are cleaning an exchanger at the beginning of the preheat train, recovery factors need to be considered. A 20°F increase at the exchanger will not gain 20°F at the heater inlet. Consider this when performing economic evaluations. For the crude or vacuum tower heat balance or heat removal issues, then target that section of the tower and clean those bundles.

Another method for determining which exchangers to clean is a rigorous model. I have seen a couple programs that model crude pre-heat trains. These programs model heat integration within your crude towers and are able to predict the inlet temperature of your heater. We do not currently utilize that at CITGO. We prefer the simple go-out-in-the-field-and-look method.A couple of points to keep in mind: When exchangers are taken offline and bypassed, understand how this will affect the crude or vacuum tower heat balance. If parameters are not set properly beforehand, the crude or vacuum tower may become upset, which may result in product quality or pressuring of the tower. We experienced one event when we almost had to abort a cleaning because taking the exchanger offline resulted in an increase in crude tower pressure. We learned from that event. Now, we make sure to predict the heat balance change and adjust tower crude rate or heater firing as necessary prior to taking a preheat exchanger offline to clean.

HERLEVICH (Marathon Petroleum Corporation)

Just to add onto the cleaning piece: At our refinery, we recently made a procedure for evaluating whether it is actually acceptable to take bundles offline or bypass equipment. This has not typically been an issue in the crude pre-heat train because they were designed with online cleaning in mind. However, our practice is to review anyway because we have found several instances where downstream equipment was not rated for the bypassing scenario MAWT (maximum allowable working temperature) or sometimes the blocked-in case MAWP (maximum allowable working pressure).

We monitor the heat exchanger performance at Marathon, in most plants, by using a simple spreadsheet-based model as well. We have standardized on a corporate-supported, macro-driven, performance monitoring spreadsheet that was developed in-house. We used these on all units in the refinery, both for heat exchange and for basic catalyst performance. We monitor the heat exchanger performance in terms of actual heat transfer coefficient (UA) compared to start-of-run clean conditions. We also track heat exchanger pressure drop through field measurement. Once again, we often employ our college co-op workforce for these activities. Next, a simple economic analysis is conducted to compare rate reductions versus fuel gas savings. Most of our plants also do a ready-to-run analysis, which that typically results in spring heat exchanger cleaning programs.

A couple of points to keep in mind: When exchangers are taken offline and bypassed, understand how this will affect the crude or vacuum tower heat balance. If parameters are not set properly beforehand, the crude or vacuum tower may become upset, which may result in product quality or pressuring of the tower. We experienced one event when we almost had to abort a cleaning because taking the exchanger offline resulted in an increase in crude tower pressure. We learned from that event. Now, we make sure to predict the heat balance change and adjust tower crude rate or heater firing as necessary prior to taking a preheat exchanger offline to clean.

The next slide shows a graph from the performance monitor. This macro-driven, Excelbased software produces graphs that the tech service engineers include in their monthly reports. The results are analyzed to estimate fouling trends. You can see the UA and the duty for this particular set of exchangers in the graph.

The last slide lists the typical configurations for our crude units. In our bigger refineries, we have complete dual trains on the heat exchange side. During the spring exchanger cleanings, we will reduce the rate on those crude units to clean one heat exchange train at a time. In the medium-sized crude units, we tend to take off the individual exchangers or smaller batteries of exchangers. Finally, we have one small refinery that idles the entire unit for heat exchanger cleaning. This process is often conducted in sympathy with work on other refinery other units.

VICTOR TAILOR (Commonwealth Engineering & Construction)

The panel covered the entire pre-heat train potential problem well. I want to add that in a revamp case, you should consider the pressure drop through the heat exchanger when calculating the desalter charge pump to make sure that it has enough head, even at a slightly higher pressure drop. This pump needs to be very generously sized. In this way, you avoid reducing crude charge for a longer period of time.

TOM GERMANY (Calumet Specialty Products Partners, L.P.)

Has anyone on the panel used online thermal shocking of exchangers to clean in lieu of taking them off mechanically?

PRIBNOW (CITGO Petroleum Corporation)

Yes, we have heard of a couple of techniques in the cold pre-heat train.

SHENKLE (Flint Hills Resources, Ltd.)

We have not.

HERLEVICH (Marathon Petroleum Corporation)

No, I am not aware of our doing that.

HAROLD EGGERT (Athlon Solutions)

On units that are being monitored for crude pre-heat trains, we have noticed that when there is an upset, power failure, or shutdown, after the shutdown and startup, there is some significant heat gain recovery. It is, again, the thermal shock. The tube shrinks at a different rate than the fouling that is on there, resulting in some of the material breaking loose. I think some of that is being employed in a cold train, too.

HERLEVICH (Marathon Petroleum Corporation)

Just one more thought to add to that. We have seen times where we take down exchangers and prep them by conducting steamout and cleanup. Then where we have put them back online without any mechanical cleaning, they do still work much better. Once you discover that this process works, it may actually become routine. So take a look at that.

DENNIS HAYNES (Nalco Champion Energy Services)

The use of a heat exchanger rating program that can look at individual exchangers and how they relate to the overall system performance is the best method to rank the exchangers based on impact to the system due to fouling condition. Look at fouling rate, and then develop a cleaning schedule.

CHRIS CLAESEN (Nalco Champion Energy Services)

An intelligent way to look at this is by the determination of the total cost of operation (TCO) for each exchanger. The exchanger TCO consists of the additional fuel-burning cost required to generate the duty lost by fouling, cleaning costs (which are the additional furnace duty when offline), and fixed costs for chemicals, equipment, and manpower. More frequent cleaning reduces the additional fuel burning cost, but also increases the cost of cleaning. Clearly, there must be an optimum time to clean an exchanger, which will give the minimum cost, and this can be calculated. The exchanger network and individual exchanger fouling rate should be routinely calculated and tracked to maintain an optimum cleaning schedule using exchanger model software, such as NALCO’s Monitor™.

GARY HAWKINS (Emerson Process Management)

Emerson Process Management offers a heat exchanger health monitoring solution that calculates the duty on both the hot and cold sides (should be the same, alerts if too much deviation) and the current value of the heat exchanger coefficient. Changes in the heat exchanger coefficient are indicative of fouling. The rate of change of fouling is also calculated, and alerts can be generated to warn of potential crude instability leading to accelerated fouling of the exchangers.

PATRICK TRUESDALE (Emerson Process Management)

One problem facing refiners gathering all of the temperature measurements around each tube-and-shell bundle in the crude pre-heat train. Often, capped test thermowells are provided for periodic measurements taken manually. This option presents a refiner with the opportunity to take advantage of wireless temperature transmitters utilized to bring these missing temperature measurements into heat exchanger health monitoring solutions.