Value of monitoring exchanger networks A rig rigor orous ous ex excchan hange gerr sim simul ulati ation on mod model el ca can n be be used used to ca calc lcul ulate ate the tru true e cos costt of fouling in crude preheat networks Laura Copeland Nalco Company
H
eat exchanger fouling has a direct impact temperature limits. Fouling always leads to on protability. Over time, fouling leads energy losses when fuel has to be increased to to higher energy consumption, higher the furnace to make up for lower crude tempera maintenance costs, reduced feed rates and tures coming from the fouled crude preheat shorter intervals between turnarounds. The rela - network. tionship between fouling and energy becomes more signicant when you consider the link What causes fouling? between additional fuel gas consumption, higher There are three main operational factors that CO2 emissions and the detrimental impact on a lead to fouling: blending crudes, the velocity renery’s energy intensity index (EII). The envi - through the process and crude quality. Changes ronment and the total cost of operation (TCO) to any of these factors can lead to a change in are negatively impacted. fouling throughout the process. Proven energy savings can be realised when When crudes are blended together, there is the the fouling of a crude unit preheat network potential for instability that can lead to fouling. exchanger can be effectively monitored. If the crudes are processed individually, the foul Monitoring will determine how fouling in a ing potential can be different than if two or more network changes with time. Crude units see the crudes are blended together. For example, a highest charge rates and the largest temperature rener could be processing a heavy, low API increase of any renery unit, 1 so the benets of a crude and have very little fouling, but when they successful monitoring and fouling control blend a light, high API crude with it they see programme can be signicant. increased fouling in their crude preheat train. This article will include a brief review of crude The introduction of another type of crude has unit heat exchanger fouling mechanisms, how caused instability in what is normally a stable fouling affects energy management costs, and crude. Preheat train monitoring can be used to potential solutions. support the rener’s decision-making process when implementing a strategy to t o prevent fouling Fouling mechanisms deposition due to the incompatibility of crudes.2 What is fouling? It is the formation of deposits Figure 1 shows the results of testing done on the in process equipment that impedes the transfer Nalco Fouling Potential Analyzer (FPA), where of heat and increases the resistance to uid ow. each crude individually has a lower fouling Several physical, operational and chemical potential than when they are blended together. factors can combine to form these deposits. Most The FPA value is the inection point of each crude preheat deposits have low thermal conducconduc - trend, and a lower FPA value equals less stabilstabil tivity and reduce heat transfer. Fouling can have ity. In this example, when crude A is blended a substantial economic impact upon a rener’s with crude B, the stability decreases and, there there-protability when it causes throughput reducreduc - fore, there is a higher potential for fouling. tions due to hydraulic limits or furnace tube Anotherr operat Anothe operational ional factor that can cause foul foul--
www.digitalrefining.com/article/1000544
PTQ Q4 2012 1
event” with a sudden lowering of throughput. A well-monitored 0.09 system can locate which e 0.08 exchangers tend to foul the c n most as a result of decreased a 0.07 b r throughput (velocity). o 0.06 FPA value decreased s = less stability b 0.05 The third operational factor a that can lead to fouling issues e 0.04 v i t throughout the renery is the a 0.03 l e quality of the incoming crude. Crude A R0.02 Crude oil can contain Crude B 0.01 Crude C asphaltenes and inorganic 0 materials that can contribute to 45 47 49 51 53 fouling in the system. FPA value Asphaltenes are the most common fouling material found Figure 1 FPA trend that shows crude blending impact on stability in the hot preheat train. They are naturally stabilised by 500 resins that prevent them from agglomerating, but asphaltenes 495 can easily agglomerate when 490 destabilised and cause fouling. 485 F ° Another type of foulant is exist , 480 T I ent debris such as sand or F 475 N sediment carried in the crude 470 oil that may be deposited when 465 stressed by heat. The deposition 460 of inorganic salts can result in 455 fouling if the renery has no 14 November 3 January 22 February 12 April desalting capability or if the desalters are not working propFigure 2 Example trend of NFIT; circle indicates where a crude change took place that led to increased fouling erly. Finally, one more potential type of fouling material is poly ing is a change in the velocity through the process. meric gums that can form if a reactive stream is Initially, most heat exchangers are designed and added to the crude oil. Figure 2 shows an exam sized to achieve a maximum heat transfer at ple plot of normalised furnace inlet temperature design throughput conditions. Often these design (NFIT). This shows how a change in incoming conditions achieve very low fouling rates due to crude to a renery could have a signicant the high velocity and proper bafe design and impact on fouling. More discussion of NFIT can spacing. As throughput changes or heat exchang- be found in the next section of this article. ers are added or redesigned, velocity changes and Cost of fouling + cost of fouling control = total the rate of fouling can also change. Particulates travelling along with the crude cost of operation have a greater potential to fall out and cause It is possible to operate a crude preheat to fouling at lower velocities. A proper monitoring achieve the lowest TCO by calculating the total programme should take into account the veloci- cost of fouling and the total cost of fouling ties of the various streams, how they are control. The costs of fouling are all related to the changing with time, and the impact on fouling extra fuel burned in the furnace due to the foul and temperatures throughout the system. Just as ing layer inhibiting heat to be transferred to the many exchanger networks may encounter a crude. As the fouling increases in the different “cleaning” effect from a sudden increase in exchangers, the crude exiting each exchanger throughput, they can also experience a “fouling leaves at colder and colder temperatures. That 0.10
2 PTQ Q4 2012
www.digitalrefining.com/article/1000544
lower temperature crude must be heated up to throughput margin or even environmental penalthe xed furnace exit temperature in order for ties from ring the furnace harder. The costs of the rener to meet their target cut points. This fouling control are the total spend the rener additional fuel due to fouling is difcult to calcu- makes to clean or keep an exchanger network late without a proper heat exchanger simulator clean. This would include the maintenance clean being run on a regular basis. ing costs, extra fuel to the furnace (if an Reners will also change the pumparound exchanger is taken off-line to clean), lost rates to manipulate the cuts in the atmospheric throughput margin (if rates are reduced to tower for maximum protability. This will also clean), antifoulant chemical (if used), cleaning add or delete heat from the preheat, but this is chemical (if used) and any other cost the rener not due to fouling. A proper monitoring absorbs when taking action to reverse or control programme will be able to distinguish the the existing fouling in the exchangers. difference between temperature losses due to The most common method of fouling control is operational changes from temperature losses to take exchangers off-line and mechanically clean due to fouling. This can be achieved by calculat - them by hydroblasting or lancing the inside and ing a NFIT using a base set of operating outside of the exchanger tube bundle. Whatever conditions. The NFIT will be equal to the actual the cleaning process, the rener should add all furnace inlet temperature (FIT) as long as the the costs associated with the cleaning to deteroperating conditions remain the same. When mine the optimum time to clean. Calculating the pumparound ow rates or temperatures change, cleaning cost is relatively easy, but knowing when the heat load to the preheat will change and to clean is the hard part. affect the FIT, causing the FIT and the calculated NFIT to be different. The difference will Solutions be the result due solely to operating changes In order to calculate the true cost of fouling, a between the base case and the current case. proper monitoring programme is critical. The Figure 3 is an example of the differences that total spend on fouling and fouling control can be seen between FIT and NFIT. The NFIT discussed in this article is the total cost of opera will show the temperature decline due to fouling, tion for the crude preheat. The optimum TCO is while the FIT will show the temperature decline the lowest combined cost of fouling and spend due to both fouling and operational changes. The on fouling control.3 Each exchanger’s contribuNFIT trend is useful to show the impact of tion to the furnace inlet temperature is changing any variable that has an effect on the maximised by cleaning exchangers at the fouling rate. The decline in temperature 480 due to fouling can be converted into lost BTUs (energy) that 440 NFIT and FIT must be made up in the furnace difference due by burning extra fuel. to operational 400 changes F Incorporating furnace efciency ° , and cost of fuel, the NFIT T I 360 F reects a cost of fouling. N An antifoulant programme 320 could be added to improve 280 (reduce) the cost of fouling. The NFIT cost of the antifoulant would not FIT 240 be a cost of fouling, but it should i l l y b e r s t r y u n e b e r p r e r a r y c h u b be considered as cost of fouling J u a r u g u n u a 1 v e m M 8 A t e m e b r 2 J v e m a A 1 J 9 p control. The only costs that o F 1 e 1 6 1 3 N N o 2 S 9 5 should be considered as fouling 2 8 costs are those costs that occur due to fouling, such as increased Figure 3 Example showing the difference between NFIT and FIT that can be furnace fuel spend, lost seen due to operational changes
www.digitalrefining.com/article/1000544
PTQ Q4 2012 3
optimum cleaning cycle frequency, and good exchanger network simulators will calculate the optimum time between cleanings. The optimum TCO is totally dependent on the fouling rate. If the unit starts to process a crude that fouls at a faster rate, there will be exchangers in the network that will need to be cleaned more often and, therefore, the optimum TCO will increase. The same holds true if the unit starts to process a crude that is less fouling in nature — the optimum TCO will decrease. By trending the optimum TCO for each data set the rener can see how much is being spent and so manage cleaning to achieve the lowest possible spend. This includes all costs: cleaning, fouling, chemical, lost production, and so on. What it does not include is a discount the rener receives for processing opportunity crudes. The rener will now be able to see the added cost of processing opportunity crude(s) and include these economics into future buying decisions. Antifoulants can also decrease the fouling rate, but they add to the cost of fouling control. The addition of antifoulants has to reduce the fouling rate enough to lower the overall TCO of the preheat to justify the added cost. By using a proper simulation model on a regular basis, the rener can evaluate the benets of using an antifoulant (or not) and will always be able to schedule the right exchanger for cleaning in time to keep the preheat operating at the lowest possi ble total cost.
Conclusion Crude preheat networks can be managed to achieve the lowest possible total cost of operation. It requires a rigorous exchanger simulation model that can normalise input data to calculate the true cost of fouling. The same model should be used to calculate the optimum cleaning cycle frequency for each exchanger in order to deter mine the true cost of fouling control. In this way, antifoulant chemistries can be evaluated based
4 PTQ Q4 2012
on the impact on the total cost of operation. Do: • Use a rigorous exchanger simulation model • Normalise the furnace inlet temperature • Base cleaning decisions on the maximum impact on the furnace inlet temperature when the TCO calculation shows it is time to clean • Evaluate discounted crude purchases based on impact to TCO • Evaluate antifoulant chemistry based on impact to TCO. Do not: • Make cleaning decisions based on individual exchanger data (no way to achieve lowest TCO) • Normalise the data based on crude ow only — can cause you to miss >75% of the operating changes that affect the furnace inlet temperature.
References 1 Worrel E, Galitsky C, Energy Efficiency Improvement and Cost Saving Opportunities for Petroleum Refineries, an ENERGY STAR Guide for Energy and Plant Managers, 2005. 2 Wiehe I A, Kennedy R J, The Oil Compatibility Model and Crude Oil Incompatibility, Energy & Fuels, 14, 56-59, 2000. 3 Mason B, McAteer G, Nalco Company, Energy Services Division (USA), Crude Preheat Energy Management Leads to Sustainable Energy Savings, Hydrocarbon Processing, 105-110, Sept 2008. Laura C Copeland is Global Industry Development Manager with Nalco in Sugar Land, Texas. She holds a BS in chemical engineering from The University of Iowa and a MBA from Northwestern University. Email:
[email protected]
LINKS More articles from the following categories: Corrosion/Fouling Control Heat Transfer Process Modelling & Simulation
www.digitalrefining.com/article/1000544
