Posted in

How to size a tubular heat exchanger for liquid-liquid applications?

If you’re a process engineer, plant manager, or operations lead needing to size a tubular heat exchanger for liquid-liquid applications, odds are you’ve stared at a spec sheet, scratched your head, and wondered if your current design will work as good as it sounds on paper. I’ve been in this spot a hundred times, both sizing exchangers for our own team’s projects and walking customers through the mess of variables that make or break a liquid-liquid unit. And let’s be real: liquid-liquid applications are way trickier than heating or cooling a single fluid—you’ve got two fluids, their viscosity, fouling, temperature swings, and not to mention, if you mess up, you’re looking at either way too big an exchanger (wasting your budget) or way too small (your production line goes down, and that’s a whole other headache). Tubular Heat Exchanger

First off, let’s cut through the jargon and get clear on what we’re even talking about here. Tubular heat exchangers for liquid-liquid are, at their core, bundles of metal tubes (usually stainless steel, copper, or titanium, depending on your fluids) tucked inside a shell. One fluid runs through the tubes, the other through the shell side, and heat transfers from the hotter one to the colder one through the tube walls. Unlike air-to-liquid exchangers or gas-to-liquid units, liquid-liquid means both fluids are dense, viscous, and way more prone to things like uneven flow distribution, fouling, and pressure drop—stuff that doesn’t matter as much when you’re moving air. That’s why sizing these wrong is way more costly.

Let’s start with the non-negotiables you can’t skip before you even open your Excel sheet. I see customers all the time skipping this step and ending up with an exchanger that’s totally mismatched. First, get your duty right. The heat duty is the total amount of energy you need to transfer between the two fluids, calculated with the basic Q = m × Cp × ΔT formula. Q is your duty in BTU per hour or watts, m is mass flow rate, Cp is specific heat capacity, and ΔT is the temperature change each fluid needs. Wait, but hold on—this is where so many people mess up: they use inlet and outlet temps but forget the log mean temperature difference (LMTD) correction factor. Because the temperatures of the two fluids change along the length of the exchanger, the average temperature difference isn’t just the difference between inlets and outlets. LMTD accounts for that, and the correction factor F adjusts for whether your exchanger is counterflow or parallel flow. Counterflow is almost always better for liquid-liquid because it gives a higher LMTD, so you don’t need as much surface area. That’s a quick win right there—don’t cheap out on going counterflow.

Next up, fluid properties. This is where liquid-liquid sizing differs from any other application. Gases are low viscosity, easy to move, but liquids—especially things like heavy oils, syrups, or wastewater with solids—are way thicker. Viscosity is huge here because it affects two things: heat transfer coefficient and pressure drop. If a fluid’s viscosity is over like 10 centipoise, you’re going to see a big drop in heat transfer on the shell side, because the fluid flows slower, and the layer near the tube wall gets stagnant, acting as an insulator. Pressure drop matters too—if you size the exchanger too tight, you’ll end up with way higher pressure drop than your pump or compressor can handle, costing you a fortune in extra energy to move the fluids. I once had a customer who sized an exchanger for a high-viscosity polymer, forgot to account for viscosity, and ended up needing a 20% bigger pump just to get the fluid moving—wasted $12k in energy a year, no joke.

Fouling is another big one for liquid-liquid. Unlike gases, which don’t leave much gunk on tubes, liquids—especially process fluids with minerals, solids, or organic compounds—will build up a layer on the tube walls over time. That adds a thermal resistance, so your exchanger has to account for that extra layer. If you don’t add a fouling factor (we usually use standard values like 0.0001 to 0.0005 ft²·°F·h/BTU, depending on the fluid), your exchanger will be underrated in a few months. I’ve seen guys skip fouling factors because they thought it was “overkill” and end up replacing their exchanger every 18 months instead of every 7 years—way more expensive in the long run.

Now, let’s talk about the actual sizing steps. Let’s keep this practical, not just textbook. First, pick your exchanger type. For liquid-liquid, two most common are fixed-tube sheet and U-tube. Fixed-tube is cheaper, easier to clean for less viscous fluids, but can’t handle as much thermal expansion. U-tube is great for high temp or pressure applications, and the U-bend lets the tube bundle expand without warping, but you can’t clean the inside of the tubes as easy. If your fluids are super fouling-prone, you’ll want a floating head exchanger—those are easy to pull the bundle out and clean, which is a lifesaver for liquid-liquid. I always recommend floating head if your process has any solids or high fouling, because the downtime to clean a fixed-tube is way more than the slight extra cost of a floating head unit.

Once you’ve got the type, it’s time to calculate the heat transfer coefficient (U-value). U is basically how well heat moves through the tube walls, and it’s a combination of the tube side coefficient, shell side coefficient, wall resistance, and fouling resistance. The formula for U is 1/U = 1/hi + Rfi + Rw + Rfo + 1/ho, where hi is tube side coefficient, Rfi is tube side fouling, Rw is wall resistance, Rfo is shell side fouling, and ho is shell side coefficient. There are correlations to calculate hi and ho, like the Dittus-Boelter for tube side and the Donohue or Kern method for shell side. Wait, the Kern method is super common for basic liquid-liquid, which is perfect if you’re not a heat transfer PhD—you can use it to get a rough U-value, then adjust for your specific fluids. Pro tip: if your shell side fluid is high viscosity, use the Sieder-Tate correlation instead of Dittus-Boelter, because it accounts for viscosity changes with temperature. That’ll give you a way more accurate U.

Next, calculate the required surface area (A) using A = Q / (U × LMTD × F). This is your number one sizing metric. If you get this right, your exchanger will do the job, but you don’t want to round down too much. I always tell customers to add a 10-15% safety factor to the surface area, not just a tiny 5%. Wait, why? Because no one’s numbers are perfect—flow rates might change, fouling will build up over time, and you might have a temporary peak in production. If you only size for exactly your current duty, you’ll end up underperforming in a year when your line ramps up. But don’t go overboard with a 30% factor—you’re just wasting money on extra metal that you’ll never use.

Now, let’s talk about common mistakes people make that blow up their sizing. First, mixing up tube side and shell side fluids. Wait, which fluid goes where? For liquid-liquid, you always want the more fouling-prone or higher viscosity fluid on the tube side, because you can clean the tubes easier. If you put a super viscous fluid on the shell side, it’s way harder to clean the bundle, and flow distribution is worse. I’ve seen a customer put a heavy crude oil on the shell side, and after 6 months, the bundle was so caked in oil they had to use a jackhammer to get it out—total disaster. That’s a silly mistake, but it happens all the time.

Another mistake: not checking pressure drop. A lot of guys focus only on heat transfer and forget that pressure drop is a big operational cost. Let’s say you size your exchanger to get the perfect U and surface area, but the pressure drop on the tube side is 20 psi, and your pump can only handle 15 psi. You’re either going to have to throttle the flow (which reduces your heat transfer) or buy a $5k pump, just because you didn’t check pressure drop during sizing. For liquid-liquid, you usually want pressure drop to be between 5 and 15 psi, depending on your pump’s budget. If you’re dealing with a high-viscosity fluid, you might have to go a bit higher, but don’t go over 20 psi—otherwise, your energy costs will eat into your profits.

What about special cases? Let’s say you have immiscible liquids, like oil and water. Liquid-liquid with immiscible fluids is different because you might have entrainment, or the two fluids don’t mix evenly. In that case, you might need baffles that are spaced closer together to improve mixing, or even a shell-and-tube with a larger shell diameter to reduce velocity. Also, if you’re heating or cooling a fluid that’s near its boiling point, you have to worry about vaporization, which messes up heat transfer. For that, you might need a larger surface area, or a different tube layout to keep velocity high enough to prevent vapor pockets from forming.

Wait, let’s make this real with an example, so you can see how all this works. Let’s say you need to cool 10,000 lb/h of heavy oil from 200°F to 120°F, using water that comes in at 80°F and needs to warm up to 110°F. First, calculate Q: m = 10,000 lb/h, Cp for oil is 0.5 BTU/lb·°F, ΔT for oil is 80°F, so Q = 10,000 × 0.5 × 80 = 400,000 BTU/h. Next, ΔT for water is 30°F, so LMTD counterflow is ((200-110) – (120-80)) / ln((90)/(40)) = (50) / ln(2.25) = ~50 / 0.8109 = ~61.66°F. Since counterflow, F is 1, so LMTD × F is ~61.66. Now, U-value: let’s say oil is on shell side (5 centipoise), water on tube side, fouling factors are 0.0002 for both. U would be around 150 BTU/ft²·h·°F for liquid-liquid, but add the fouling, so adjust to 120. Then required surface area A = 400,000 / (120 × 61.66) = ~54 ft². Add 10% safety factor, so ~59 ft². Now, match that to a tubular exchanger—say, a 1-shell pass, 2-tube pass fixed-tube exchanger with ¾ inch tubes, that gives you ~60 ft². That’s your size. If we’d done parallel flow, LMTD would be lower, like ~45°F, so we’d need ~88 ft²—way bigger, so counterflow is the clear choice here.

Now, as someone who’s been selling tubular exchangers for years, I’ll tell you what the process usually looks like after sizing. Most of our customers don’t have all the exact numbers off the bat—flow rates are a little off, fluid properties might change, they’re not sure about fouling factors. That’s why we don’t just spit out a number. We work with you to adjust, because we know every application is different. For example, if you’re working with food and beverage liquid-liquid, like mixing chocolate and milk, you need sanitary tubes, easy cleaning, and a U-value that’s higher than industrial fluids—we can tweak the tube size and material to get that. If you’re working with chemical processing, where fluids are corrosive, we use titanium tubes instead of stainless, and adjust the wall thickness to handle pressure.

Wait, another thing I wish more customers knew: sizing isn’t a one-and-done. If you change your process later, like increasing flow rate by 20%, you might need a slightly bigger exchanger, or you can adjust the baffle spacing or tube passes to get a better U-value. We do retrofits all the time—customers who have an old exchanger that’s underperforming, and we size a new one or modify the existing one to fit their new needs. That’s way cheaper than tearing out the whole line, too.

Let’s wrap up with the key takeaways so you don’t forget all this: 1. Get your duty right, don’t skip LMTD and F factors—parallel flow is for losers here, go counterflow. 2. Account for fluid properties: viscosity kills heat transfer, so size for that, not just “water is easy”. 3. Fouling factor is non-negotiable—skip it, and you’ll be replacing your exchanger faster. 4. Pick the right exchanger type: floating head for fouling, U-tube for high temp/pressure, fixed-tube for low cost and easy cleaning. 5. Add a 10-15% safety factor, not more, not less. 6. Always check pressure drop—an exchanger that fits your heat duty is useless if your pump can’t move the fluids through it.

If you’re still stuck on sizing, or you’ve got a specific liquid-liquid application (even the weird ones, like high-viscosity slurries or immiscible fluids) that you need to get right, don’t guess. We help process engineers and plant managers every day get the right tubular heat exchanger sized, installed, and working without the headaches of being under or over spec. Just reach out, tell me your flow rates, temps, fluids, and I can walk you through what will work for your budget and operations.

Floating Head Heat Exchanger References

  1. Kern, D. Q. (1950). Process Heat Transfer. McGraw-Hill.
  2. Kakac, S., & Liu, H. (2002). Heat Exchangers: Selection, Rating, and Thermal Design. CRC Press.
  3. Standards and Practices for Heat Exchangers, Tubular Exchanger Manufacturers Association (TEMA), 10th Edition, 2019.
  4. Sieder, E. N., & Tate, G. E. (1936). Heat Transfer and Pressure Drop of Liquids in Tubes. Industrial & Engineering Chemistry, 28(12), 1429-1435.

Beijing Cycle Columbus Environmental Science & Technology Co., Ltd.
Beijing Cycle Columbus Environmental Science & Technology Co., Ltd. is one of the most professional tubular heat exchanger manufacturers and suppliers in China. We are committed to providing high quality customized products with competitive price. Please feel free to buy advanced tubular heat exchanger made in China here from our factory. Also, quotation is available.
Address: 1201, 12/F, Tower B, Bldg 8, Guancheng Garden, Haidian District, Beijing, China
E-mail: Info@saikekanglun.com
WebSite: https://www.cycle-columbus.com/