Advanced Shell-and-Tube Longitudinal Flow Technology for Improved Performances in the Process Industry

Abstract

Several heat exchanger technologies have been developed in the second half of the former century thenceforth for addressing a multiplicity of incumbent topics shaping the discussion in the technical community and the economics of the process industry. In the frame of shell-and-tube layout, longitudinal flow deserves a peculiar place. Initially conceived for addressing requisition for reduced vibration and fouling accumulation and later recommended in case of limited allowable pressure drops, it proved valuable in replacing segmental layouts whereas weight and footprint come into the picture and reliability matters. Structural increases in the cost of raw materials and expectation for extended operational continuity push the industry in the direction of more efficient and dependable technologies. This chapter focuses on the EMbaffle® design, among the most reputed longitudinal flow shell-and-tube technologies, whose extensive adoption in oil and gas, chemical sector, and renewable power generation in the last two decades allows some fair yet not exhaustive considerations. After a concise introduction to the feature of longitudinal flow technology and to EMbaffle® basic design equations, measures of performance will be discussed. Comparison with conventional technologies will be outlined. Selected realizations will be critically presented and their potential for effective market penetration duly assessed.

1. Introduction

In the second half of the twentieth century, the impressive growth of the process industry, triggered by broad economic development, promoted the emergence of new technologies. Process engineers started thinking in terms of innovative designs capable of addressing issues whose frequency could not be neglected in view of demanding plant performances.

Shell & tube (S&T) heat exchangers did not make an exception. As the workhorse in the industry, they provided different services under diverse temperature and pressure operating conditions and fluids, emerging as a reliable solution where reliability and performance meet. Double and triple segmental layouts and TEMA types (i.e. G, H, J and X), designed for addressing lower pressure drop requirements by saving at the same time the advantages of the robust and rugged S&T layout, added to well-established single segmental layouts based on E and F TEMA types [1, 2, 3]. While standard design proved robust and reliable and guaranteed safe operations in critical conditions and even harsh environments for decades, weak performances could occur with negative impacts on maintenance costs and exchanger life; yet consequences in terms of line shutdown due to equipment failure could hardly be underestimated. Novel designs aimed to overcome limitations of conventional designs (i.e., segmental), the ones that suffered with crude oil and high gravity residues are often rich in impurities, sometimes of a sticky nature; yet larger flow rates driven by increased production levels, solicited inception of vibrations which could damage the tube bundles. Moving from the existing design characterized by baffles placed normally to the direction of the flow and inducing cross shell-side flow, new baffles were conceived and tested in a variety of services, founding their specific field of application at the end.

The case of No Tube in Window (NTIW) embeds a peripheral shell-side chamber to streamline the cross flow toward a quasi-longitudinal flow and found wide application, whereas large flow rates require minimization of cross components of velocity impacting on the bundle. Since the tubes in the peripheral chamber cannot be properly supported, they would be most susceptible of vibration, and therefore are suppressed, such as all the tubes pass through all the baffles (Figure 1).

Figure 1.

NTIWs layout.

Disk & Doughnut, by replacing double segmental baffle with rounded baffles shaped as alternate disks and annulus, increases the longitudinal versus transversal contribution of the flow reducing the pressure drops. So, they are still widely used, whereas pressure drops are a limiting factor (Figure 2). Twisted tubes^® work on tube geometry by “twisting” the tubes around their axis, therefore inducing a quasi-longitudinal shell-side flow under turbulent profile. This baffles-free layout is profitably used whereas fouled fluids dictate low accumulation to avoid of deposits on the heat transfer surfaces [4]. Finally, Helixchanger^® emerged as a disruptive design for addressing heavy fouled fluids thanks to inclined segmental baffle geometry which favors processing of the same under reduced deposition. It was proved in several services featuring viscous and dense fluids [5].

Figure 2.

Disk & Doughnut layout.

The above solutions were developed as improved conventional design within the frame of the conservative oil and gas world.

In the 1970s, a team of engineers of Phillips Petroleum Co., while investigating the damage that occurred to a waste heat recovery unit due to induced fluid-dynamics vibrations, suggested changing the existing baffle layout and moving to a radically new one. By replacing the same with rods, initially arranged over elliptical-shaped disks and later over circular shapes, as it became customary, the Rod Baffle design (Figure 3) was on the air [6, 7]. The initial aim had been to find a promising solution for vibration issues but, due to the pure longitudinal flow layout which clears vibration at its root, it emerged for addressing a multiplicity of topics, becoming a standard in the industry, especially in the North American world.

Figure 3.

Rod Baffle layout.

Rod Baffle triggered new thinking that eventually resulted in the flourishing of designs in recent decades. Replacement of rods with strips and with structures capable of increasing the confinement of the tubes just demonstrated the vitality of the original concept, and here is where EMbaffle^® originated.

EMbaffle^® is based on an expanded metal grid type of baffle, fully supporting the tubes and improving heat transfer while containing pressure drops (Figure 4). Conceived and initially developed by Shell Corporation at the outset of the new millennium for addressing the fouling issues experienced in the Group Refineries, the technology, owing to the continuous developments by Brembana&Rolle (B&R), found promising fields of application due to benefits of longitudinal design jointed with patented open structure feature. The next chapter will explore the benefits and wide applicability of longitudinal flow based on peculiar technical features [8, 9, 10].

Figure 4.

EMBaffle^® layout.

2. Longitudinal-type heat exchanger design

Longitudinal flow layout, i.e., establishment of parallel flow between shell-side and tube-side fluids, allows pure co/counter-current relative flow to be assessed and results in a minimum temperature approach (augmenting the efficiency of the unit); it promotes the reduction of required heat transfer area, footprint, and weight. Also, parallel flow allows avoidance of cross-components of the velocity of shell-side fluid on the tube bundle, inception of vibrations, extended maintenance frequency, and improved unit reliability.

Yet, the less tortuous and shorter path followed by the shell-side fluid with respect to cross flow configuration produces lower pressure drops which favor routing of larger mass flow rates, improving the performance of existing units in plant modernization projects (extending the operating windows). The above is at the basis of lower fouling accumulation in longitudinal flow-based designs, being suppressed by any hurdle, which would cause low velocity and recirculation regions where sticky fouling might grab and adhere.

It must be remarked that longitudinal flow is not by itself likely to trigger effective heat transfer unless promotion mechanisms are put in place; here is where the different technologies differ and where the proprietary concept generally lies.

Bundle construction follows the consolidated conventional baffles knowledge, simply replacing the same with Rod style baffles or other engineered solutions.

Although the operation of the units does not pose specific issues, maintenance may greatly be advantaged either in terms of reduced frequency of maintenance or occurrence of bottlenecking; cleaning of the bundle is generally achieved with the usual high-pressure water-jet methodology.

Rod Baffle paved the way for the deployment of several longitudinal flow designs, all sharing the above benefits.

The open geometry of tube supporting elements fits for large flow rates. Moreover. installing a vapor belt, i.e., a peripheral ring with slots, downstream the inlet nozzle through homogenization of the flow ensures prompt establishment of longitudinal flow.

Thinking that Rod Baffle has been primarily conceived for supporting the tube bundle in order to prevent vibration-induced tube damage, it’s no surprise that the performance in terms of heat transfer was not focused on by the designer, further missing the advantage of the low pressure drops attainment; so Rod Baffle resulted to be the equipment of choice specifically for addressing vibration issues.

So the potential of longitudinal flow had still to be exploited well beyond Rod Baffle; the ensuing deployments will be addressed in the next chapters.

The above considerations permit us to sketch the applicability frame of the longitudinal flow layouts. Firstly, the process engineer will consider their adoption in all the cases’ efficiency, effectiveness, or operating window, as well as large mass flow rates are involved. Secondly, by aggregating the technical features, it is possible to deduct the fields of application quite straightforwardly: large flow rates under low pressure drops along longitudinal flow suggest applicability to a gas service, while longitudinal flow jointly with low-temperature approach encompasses a generalized feed-effluent service; longitudinal flow under contained pressure drop is typical for a fouled service and so forth. The list includes conventional and emerging gas services, from gas pre/inter/after cooling in compression stations to CO₂ utilization in advanced power cycles, from gas dehydration in offshore platforms to synthesis of hydrogen, in the fertilizing industry along the nitrogen production chain and in more general gas–gas interchanger applications.

3. EMbaffle^® technology

Diamond-shaped grid characterizes the geometry of EMbaffle^® design and governs its thermo-hydraulic behavior.

Patented know-how based on so-called expanded metal production process, allows the design of different grid shapes for any tube diameter of practical interest.

Fluid dynamics of flow across a grid structure has been well known since the 1990s [11, 12, 13]; experimental studies conducted in laminar and turbulent regimes describe the grid as a turbulence promoter by (i) destroying the laminar regime and (ii) superimposing additional turbulent modes to the existing flow structure. The three-dimensional nature of the added modes can be expressed in terms of a single parameter, named turbulence intensity, which shows a prominent peak in the proximity of the grid. While a simplified theoretical formulation of such parameter, which allows its correlation with the geometry of the grid is still not available, CFD tools allow it to be represented conveniently as a function of related Reynolds and Prandtl numbers.

Fluid-dynamics is more complex whereas tubes are inserted into the diamond-shaped grid; flow is pushed inside the grid by the driving pressure force and expands downstream. Turbulent modes generated during the expansion of the fluid stream after the grid add to the modes generated in no-tube configuration; expansion cooperates in increasing turbulence through the well-known expansion-cone effect. This depends on the angle the grid is inclined, and it achieves a maximum in correspondence with a narrow range of slopes.

Proprietary design tools, supported by CFD and experimental validation process, permit to qualify each single grid geometry in terms of intensity of turbulence for specified Reynolds number. Being heat transfer coefficient and pressure drops directly related to the turbulent intensity, it follows that, for assigned process conditions, the EMbaffle^® exchanger may be customized in terms of required duty over allowable pressure drops by selecting the grid shape and its number for a specified tube diameter (Figure 5).

Figure 5.

Turbulence versus Duty. Governing turbulence for customizing heat transfer in EMbaffle^® (illustration rights by B&R).

The ability to shape the geometry of the grid for increasing the shell-side heat transfer and accommodating the desired number of tubes adds a valuable degree of optimization. Yet the open structure of the grid geometry plays a role in preventing fouling accumulation typical of segmental baffles, which ultimately explains the reasons of the early fortune of the technology.

The University of Cambridge [14] developed the SmartPM tool for evaluating fouling deposition in S&T heat exchangers, actually integrated in the HTRI platform [15]; EMbaffle^® resulted in very low fouling accumulation over several Refinery services.

In the early 2000s, a major effort was put into play in order to identify the governing equations for EMbaffle^® technology. Moving from Rod Baffle’s established equations for shell-side heat transfer coefficient and pressure drops, experimental tests, conducted in recognized third-party test bench facilities, permitted to finalize the grid-dependent coefficients for applicable thermo-hydraulic equations, either for laminar and turbulent regimes, in the Re range 1 × 10² to 1 × 10⁵.

Heat transfer correlations for laminar and turbulent flow are respectively:

$$\mathrm{Nu}=C_{\mathrm{L}}{{Re}_{\mathrm{h}}}^{0.6}{Pr}^{0.4}{\left(\frac{{\mu }_{\mathrm{b}}}{{\mu }_{\mathrm{w}}}\right)}^{0.14}$$E1

$$\mathrm{Nu}=C_{\mathrm{T}}{{Re}_{\mathrm{h}}}^{0.8}{Pr}^{0.4}{\left(\frac{{\mu }_{\mathrm{b}}}{{\mu }_{\mathrm{w}}}\right)}^{0.14}$$E2

The geometry coefficient functions, $C_{\mathrm{L}}$ and $C_{\mathrm{T}}$, account for the enhancement due to the cross flow at the shell entrance and exit and by-pass flow. The Reynolds number is calculated as:

$${Re}_{\mathrm{h}}=\frac{\rho U_{\mathrm{S}}{D}_{\mathrm{h}}}{{\mu }_{\mathrm{b}}}$$E3

where $U_{\mathrm{S}}$ is the shell-side velocity and $D_{\mathrm{h}}$ is the thermal characteristic diameter.

The shell-side velocity is calculated with the continuity equation, using the following expression for the shell-side flow area:

$$A_{\mathrm{s}}=\frac{\pi }{4}\left({D_{\mathrm{s}}}^2-{N_{\mathrm{T}}{D}_{\mathrm{T}}}^2\right)$$E4

The thermal characteristic diameter is calculated as:

$$D_{\mathrm{h}}=\frac{4\left(P_{\mathrm{T}}-\frac{\pi }{4}{D_{\mathrm{T}}}^2\right)}{\pi D_{\mathrm{T}}}$$E5

Experimental measures permitted to validate the general heat transfer correlations and the coefficient $C_{\mathrm{L}}$ and $C_{\mathrm{T}}$ for different grid geometries. In Figure 6 the measured Nusselt number as a function of Reynolds number is reported. The kink in the prediction curve sets the transition between laminar and turbulent regimes.

Figure 6.

Measured Nusselt number as a function of Reynolds number.

Pressure drops are calculated as the sum of the longitudinal flow component and the baffle flow component:

$$∆P=∆P_{\mathrm{L}}+∆P_{\mathrm{B}}$$E6

The expression for the longitudinal component is:

$$∆P_{\mathrm{L}}=\frac{2\rho f_{\mathrm{F}}{L}_i{U_{\mathrm{S}}}^2}{D_{\mathrm{P}}}$$E7

where$D_{\mathrm{P}}$ is the hydraulic characteristic diameter, $f_{\mathrm{F}}$ the Fanning friction factor and $L_{\mathrm{T}}$ the length of the tubes.

$D_{\mathrm{P}}$ is calculated as follows:

$$D_{\mathrm{P}}=\frac{4\left[\frac{\pi }{4}\left({D_{\mathrm{s}}}^2-{N_{\mathrm{T}}{D}_{\mathrm{T}}}^2\right)\right]}{\pi \left(D_{\mathrm{s}}+N_{\mathrm{T}}{D}_{\mathrm{T}}\right)}$$E8

The friction factor is based on the following expression:

$$f_{\mathrm{F}}=\left\{\begin{array}{c}\frac{16}{{Re}_{\mathrm{p}}},\text{laminar}{Re}_{\mathrm{p}}\\ \frac{0.079}{{{Re}_{\mathrm{p}}}^{0.25}},\text{turbulent}{Re}_{\mathrm{p}}\end{array}\right.$$E9

The baffle pressure drop is calculated using the baffle velocity $U_{\mathrm{B}}$ and a baffle loss coefficient $K_{\mathrm{B}}$:

$$∆P_{\mathrm{B}}=K_{\mathrm{B}}{N}_{\mathrm{B}}\frac{\rho {U_{\mathrm{B}}}^2}{2}$$E10

where $N_{\mathrm{B}}$is the number of the baffles.

The baffle velocity is determined using the continuity equation with the following definition of the baffle flow area:

$$A_{\mathrm{B}}=A_{\mathrm{S}}-A_{\mathrm{R}}-A_{\mathrm{EM}}$$E11

$A_{\mathrm{R}}$is the ring area, while $A_{\mathrm{EM}}$ is the projected area of the EMbaffle^® grid upon the plane normal to the flow direction and depending by the grid geometry.

$K_{\mathrm{B}}$ is the correlation factor accounting for the effect of grid porosity and degree of establishment of longitudinal flow inside the unit, depending on the ratio $A_{\mathrm{B}}/A_{\mathrm{S}}$ and from the shell length to diameter ratio.

Experimental measures permitted to validate the general pressure drops correlations for different grid typologies. While correlations resulted in good agreement with measured data for high Re numbers and low viscosity fluids (i.e., for typical gas services) and for mid Re numbers (Figure 7), prediction at low Re numbers for viscous fluids, resulted less accurate, possibly suggesting the lower attitude of the grid as turbulence promoter while enveloped by a heavy viscous stream.

Figure 7.

Pressure drops as a function of Reynolds number for mid Re.

Indeed test-bench conducted in real Refinery environment in late 2004 permitted to assess lower fouling accumulation of same crude oil stream upon EMbaffle^® unit, with respect to segmental unit, whereas both were placed in parallel on the same line. Further research performed by independent third party witnessed a potential for fouling reduction of five to seven times, in terms of deposited thickness over the same time frame. The consequences in terms of reduced maintenance and improved reliability descend straightforwardly.

The above anticipates some conclusions among applicable technologies. Longitudinal flow designs configure themselves as consistent alternative, whereas limited pressure drops, vibration issues, and fouling accumulation may deteriorate the performances of the unit; in this sense, their application directly descends from process requirements.

On the other side, as it is emerging with growing awareness, customization is a strong drive in the adoption of these technologies. In the case of EMbaffle^®, the ability to govern fluid-dynamics by selecting the grid geometry and the number of baffles, permits to work on turbulence generated across the single baffles and the heat transferred as a consequence of it. Accordingly, EMbaffle^® may be basically fitted-for-purpose, to an extent not achievable with segmental flow technologies, in which compromise between the thermal performance and the corresponding pressure drop has always to be searched.

Customization can be implemented for (i) reducing temperature approach, i.e., the active surface of the unit, then the dimensions and weight of the same, in some cases even the number of shells required for service; (ii) reducing bundle-related pressure drops, i.e., reducing the diameter of the unit and the thickness of the shell (especially valuable in case of high pressure services); (iii) optimizing the ratio of the heat transferred per units of pressure drops, by allowing optimized unit shaping for siting and power saving of routing pumps, compressors, and fans.

Tables 1 and 2 sum up the benefits categories of user experience in the adoption of the technology. Table 1 concisely reports the key spillovers produced by the selected performance criteria; Table 2 reports the impact exerted on the investments (CAPex) and O&M costs (OPex) and the relevance of the latter for the main stakeholders involved in the deal.

Measures of performance	Technological spillovers
Efficiency (temperature approach)	1. Compact dimensions 2. Reduced weight 3. Critical temperature services (i.e., close to freezing/vaporization points or tight-range operational points)
Effectiveness (heat transferred per units of pressure drops)	4. Optimized shape 5. Reduced size and power consumption of pumps/compressors
Operating window (pressure drops)	6. Same unit for increased mass flow rates 7. Thinner shell/s under reduced size 8. Optimized maintenance of parallel lines
Reliability	9. Minimized vibrations 10. Reduced fouling deposition and accumulation 11. Reduced erosion of tubes bundle

Table 1.

Measures of performance.

	CAPex	OPex	End user investment dept	End user field and maintenance	EPC	Fabricator
1.	✓		✓		✓	✓
2.	✓		✓	✓	✓	✓
3.		✓		✓
4.		✓		✓
5.		✓		✓
6.	✓	✓	✓	✓	✓
7.	✓		✓		✓
8.	✓		✓		✓	✓
9.		✓		✓
10.	✓	✓	✓	✓
11.	✓	✓	✓	✓
12.	✓	✓	✓	✓

Table 2.

Investment and operational costs perspective.

End users and EPCs represent the usual key decision players, hence ability to design-to-efficiency, -effectiveness, and other measures of performance allows the full valorization of the technology.

In this regard, comparison with reference conventional technology is suggested to the process engineer in order to promptly evaluate the convenience of alternate approach, based on the specific constraints raised from the field. Availability of EMbaffle^® technology on HTRI design platform facilitates the preliminary job and may provide some useful insights over customization range for the specific service.

4. Applications and design cases

4.1 Gas dehydration

Wet gas, as extracted from subsea basin, may contain a relevant amount of water; part of it results in being bonded to the hydrocarbon blend and cannot be effectively removed through physical separation only. At pressures and temperatures well above the expected, water solidifies trapping the hydrocarbon inside a cage-like structure, known as hydrate. Hydrates may obstruct the section of the transportation pipeline and cause severe damage to the line and the equipment. One way to face this issue is to inject an effective inhibitor like mono-ethylene glycol (MEG) whose action is to absorb the water vapor.

Heat exchangers for gas dehydration service feature a MEG inlet manifold equipped with MEG sprayers at their ends for uniform distribution of the inhibiting agent inside the tubes where wet gas is routed and water content progressively removed. Figure 8 illustrates typical spray particle distribution in the channel of the heat exchanger just upstream the tubes inlet. Heat for process completion is supplied by gas already dried, hence a feed-effluent design is in play.

Figure 8.

Particle mass distribution of inhibiting agent. (Illustration rights by B&R.)

Placed close to the gas extraction section, extremely high inlet gas pressures are common and content of vapors dictate adoption of suitable materials, being martensitic steels frequently used.

Conventional design for feed-effluent service and gas dehydration specifically is NTIW. Motivations rely on proved robustness and reliability; on the other side, extended dimensions and inherent large pressure drops makes it less viable, whereas offshore platforms a/o floating vessels host the gas treatment facility and space and weight become critical factors.

EMbaffle^® proves generally lighter than NTIW and better exploiting the allowable space while satisfying the process requirements (Table 3).

Item	Conventional design	Alternate design	Units
TEMA type	NEN	NEN	—
Number of equipment	3 parallel × 1 series	3 parallel × 1 series	—
Shell ID	1500	1350	mm
Tube length	13,000	13,000	mm
Baffle arrangement	NTIW	EMbaffle®	—
Installed area	14,589	13,740	m^2
SS pressure drops	1.0	1.0	bar
Duty	17,700	17,700	kW
Duty/installed area	1.213	1.288	kW/m^2
Weight	338.4	228.4	tons

Table 3.

Gas dehydration.

Furthermore, it proves valuable in reducing the unit length and mainly the shell diameter (in consideration of high pressures and thickness of the same).

4.2 Synthesis gas loop in ammonia production

Process heat exchangers (PHE) are used in synthesis gas loop in ammonia production, The Haber-Bosch process achieves conversion of hydrogen to ammonia in a catalyst-based converter; reaction rate is around 30%, hence multi-pass conversion suiting the loop-style approach is required.

Synthesis gas coming out from the Ammonia converter and rich in Ammonia is cooled to favor the subsequent separation of liquid Ammonia from unreacted hydrogen gas which is re-routed to the converter.

Large amount of heat carried by the synthesis gas after conversion is recovered by PHE.

No doubt that hydrogen conversion temperatures and pressures (typically 400–45°C at 140–220 bar) solicit attention to activation of corrosion mechanism, being hydrogen embrittlement, nitriding and hydrogen stress cracking the most recurrent ones.

Conventional, i.e., horizontal, layout of PHE is based on natural convection unit with separated steam drum on top of it and risers and downcomers connecting the two of them.

While keeping natural convection mechanism, i.e., natural flow of water to steam driven by buoyancy and its superior reliability, vertical U-tubes solutions had been developed aiming to a more compact design. In this regard, EMbaffle^® offers an extremely simplified design with integrated inner steam drum at top of the shell.

U-tubes are arranged according to fountain lay-out; extensive use of low alloy grades in tube construction and ferrules in austenitic material for tube inlet prevent corrosion inducement.

Shell side boiling water, heated up by the tube-side synthesis gas, rises to the steam drum where liquid droplets are separated from vapor, which is generally routed, eventually following superheating step, to a utility line (for direct utilization or power production).

Due to open grid structure, EMbaffle^® fits perfectly allowing full unconstrained cross and parallel free flow all along the vertical tube arrangement, promoting the homogeneously distributed density of boiling water/steam at any shell section level, with suppressed turbulence. The resulting extremely low pressure drops promote the highest recirculation factors, driving to increased conversion rates with reduced installed surface. On a performance basis, construction is the most simplified and cheap solution available in the market.

Innovation, in course of patenting, results quite simple as compared with generally complex alternate proprietary lay-outs and favors lean construction and assembly.

Figure 9 illustrates a 5 years’ operating unit in North America ammonia plant.

Figure 9.

EMbaffle^® arrangement for process gas boiler.

Advantages in terms of reduced footprint, weight, and complexity emerge: savings to CAPex (due to compact size and lighter weight) and OPex (impact of weight upon transportation and siting, while reduced complexity means higher reliability) do follow.

4.3 Nitric acid production

Nitric acid is a main precursor in the production of inorganic fertilizers of commercial cut; Ostwald process achieves the oxidation of ammonia in a catalytic reactor. Reactor releases nitrogen monoxide which is cooled in a heat exchanger line (HEL) and further oxidized to nitrogen dioxide before being routed to the absorption column where is converted in nitric acid by addicting water. Unconverted nitrogen dioxide and impurities (a blend referred as tail gas) are routed back to the HEL to cool the nitrous gas.

Materials of construction are challenging in consideration of corrosive attack of nitrous gas, especially at tube inlet, where it comes below its dew point: austenitic 18/10 grades may be therefore prescribed for tube and even shell construction.

HEL is a quasi-feed effluent exchanger line, made of multiple units in series, known as tail gas preheaters: Rod Baffle layout is a consolidated design, as common in any feed effluent requests (under limited pressure drops and no request for additional input of heat). Request for improved competitiveness move the process engineer to identify technological solutions improving overall effectiveness by keeping the advantages of longitudinal flow layout.

EMbaffle^® proves valuable in reducing the unit length and, mostly, the shell diameter (in consideration of high pressures required and thinner thickness of the same), under same process prescriptions. Reduction in weight and footprint further facilitate transportation of the unit/s to site and siting in existing facilities.

Advantages in terms of reduced footprint, weight, and complexity emerge: savings to CAPex (due to thinner thickness, more compact size, and lighter weight) and OPex (reduced complexity means higher reliability and low maintenance demand) do follow.

Table 4 provides comparison against conventional (Rod Baffle) layout under same process requirements.

Iitem	Conventional design	Alternate design	Units
TEMA type	NEN	NEN	—
Number of equipment	1 parallel × 1 series	1 parallel × 1 series	—
Shell ID	2080	1745	mm
Tube length	8700	7030	mm
Baffle arrangement	Rod Baffle	EMbaffle®	—
Installed area	2544	1819	m^2
Overall pressure drops	9.0	9.0	kPa
Duty	8373	8373	kW
Duty/installed area	3.291	4.603	kW/m^2
Weight	60.9	39.4	tons

Table 4.

Nitric acid production.

4.4 Naphtha hydrotreating

Heavy naphtha must be (hydro)treated before being reformed in the catalytic unit to remove sulfur, hydrogen a/o metals. Lighter fractions are treated too, before being cracked and reduced to shorter chains for producing distillate blends (e.g., diesel fuel, kerosene or even gasoline further to catalytic reforming).

Role of hydro-treating in refinery is therefore hard to overestimate.

Hydro-desulphurization (HDS) is a major class of hydro-treating processes, aiming at reducing the sulfur content in the naphtha stream before it is routed to the catalytic reformer: scope is to preserve the catalyst from poisoning, produce commercial naphtha cut and reduce the environmental impact of the same.

Separation of sulfidic acid from the hydrocarbon chain occurs in the converter at 280–420°C at high partial pressures of hydrogen; heat recovery units are required to recover the heat in the sulfur-free naphtha effluent and supply it to the fresh charge, in a typical feed effluent layout.

Materials of construction are critical due to hydrogen sulfide combined with hydrogen at high temperature and partial pressure: low alloy steels based on molybdenum and chromium are widely used to face thermal creep and hydrogen embrittlement; yet, whereas design temperatures are lower carbon steel may be used.

Segmental S&T exchangers are used in consideration of their robustness and reliability; yet, in case of modernization projects addressing increase in processed mass flow rates on both sides, whereas limitations exist for accommodation of novel units, E to F TEMA layout shift may provide a valuable hint to designer. Shift shell is checked against additional pressure drops which may exceed the limitations of the segmental unit, as it does occur frequently.

Longitudinal flow designs provide one additional drive to adoption of F layout. Equivalence of pressure levels across the longitudinal baffle, due to very limited pressure drops along the exchangers, favors its adoption; benefits range from reduced unit length to easier accommodation of novel units in existing areas.

Table 5 provides comparison of EMbaffle^® against conventional (single segmental) solution under same process requirements; need for extended exchanger line length penalizes the pressure drops, the number of units and the overall footprint.

Item	Conventional design	Alternate design	Units
TEMA type	AES	AFS	—
Number of equipment	2 parallel × 8 series	2 parallel × 4 series	—
Shell ID	660	1067	mm
Tube length	9144	9144	mm
Baffle arrangement	Single segmental	EMbaffle®	—
Installed area	4450	3791	m^2
SS pressure drops	9.32	1.90	bar
Duty	68,000	68,000	kW
Duty/installed area	15.28	17.94	kW/m^2
Weight	—	—	tons

Table 5.

Naphtha HDS.

4.5 Thermal storage in renewable power plants

In the first decade of the twenty-first century, concentrated solar power plants paved the way toward a new approach in utility scale power production. The concept is that solar heat radiation can be gathered and used for producing high temperature water steam for power production. A suitable intermediate fluid (HTF) carries the solar heat and releases it to water for producing steam. The drawback is that sun is not available throughout the entire day and during cloudy periods; more-over consumption loads may require some flexibility in power delivery.

In order to decouple and store the energy generated by the plant, in the so called Parabolic Through layout, process engineers conceived a smart system made of two large tanks filled with molten salts, selected as preferred thermal storage material. During the sunny day hot HTF, exceeding the production request, heats up the molten salt stored in the cold tank and routed to the hot tank; during late day hours and in cloudy days hot molten salt heats up the HTF before being routed to the cold tank.

A set of heat exchangers transfer heat between the molten salts and the HTF in a daily cycling process. Early technology was based on available segmental units; double segmental was preferred due to lower pressure drops under the typical large mass flow rates of molten salts, generally routed on shell side, also to simplify recovery in case of salt freezing during the night.

The solution proved robust but expensive: for standard European 50 MW plant size, six large exchangers were required to accomplish the duty.

Replacement of double segmental with EMbaffle^® permitted to halve this figure (Figure 10), by leveraging on its superior effectiveness, i.e., same heat transferred over lower pressure drops. Technology, proved in several plant configurations in Europe, North and South Africa, demonstrated its adaptability to different requirements in terms of transportation to site, siting constraints and logistic requests.

Figure 10.

EMbaffle^® units in thermal storage service in a CSP facility.

Advantages in terms of reduced number of units and complexity emerge: savings to CAPex (due to reduced weight) and OPex (due to impact of weight upon transportation and siting, while reduced complexity means higher reliability) do follow.

Table 6 provides comparison of EMbaffle^® against better conventional (double segmental) solution under same process requirements.

Item	Conventional design	Alternate design	Units
TEMA type	BFU	BFT	—
Number of equipment	6 series	3 series	—
Shell ID	2860	2130	mm
Tube length	10,400	21,000	mm
Baffle arrangement	Double segmental	EMbaffle	—
Installed area	27,646	22,025	m^2
SS pressure drops	3.0	2.1	bar
Duty	134,580	134,580	kW
Duty/installed area	4.87	6.11	kW/m^2
Weight	810	595	tons

Table 6.

Thermal storage in CSP plants.

5. Conclusions

Growth in demand of energy solicits development of new technologies, compliant with tighter environmental friendly requirements and motivated by awareness of progressive scarcity of fossil fuel resources. S&T heat exchangers, among the draft horses in the industry, have always been part of this process. Longitudinal flow technology bloomed in the second half of the last century to address specific issues, later becoming a comprehensive solution in view of its customization-ability, whereas conventional (segmental) design failed or simply resulted out of the balance. EMbaffle^® technology emerged as a valuable option, addressing diverse process engineering and field requests in green-field projects and even brown-field replacements of existing units. Technology may provide lower pressure drops and higher heat transferred per unit length of the exchanger, larger mass flow rates for same pressure drops, tighter temperature approaches, reduced fouling accumulation, and no-inception of vibration on the bundle. Yet, these benefits translate in terms of more compact sizing, reduced footprint, and weight with positive impacts on CAPex and OPex. It’s not surprising that several services in the gas, petrochemical, and chemical industries took profit from it. Feed effluent services did stand out as prominent application field: in the gas treatment and purification sector, EMbaffle^® proves limited pressure drops drive the choice of the technology, whereas in the catalytic reforming sector, the driver is the high partial pressure of the hydrogen stream which recommends the containment of the unit diameter. Finally, in the chemical and fertilizing industry, large mass flow rates of feed charges make full bore technology largely preferred over alternatives. Moving from historical referrals in crude oil preheating and Over’d condensation, EMbaffle^® waded in the gas sector supplying large inter-cooling units for on-shore compression pipelines and later off-shore compression facilities aboard FLNG vessels. Confidence achieved in these developments triggered innovative concepts, actually embedded in standard designs. Working closely in touch with Process Licensors and Engineering Co., novel solutions are identified, and improvements are elaborated on solid bases. Upon these premises, EMbaffle^® technology will continue to pave a remarkable way in the industry.

Nomenclature

$A$	heat transfer area (m^2)
$A_{\mathrm{B}}$	Baffle flow area (m^2)
$A_{i\mathrm{M}}$	EMbaffle grid projected area (m^2)
$A_{\mathrm{R}}$	ring area (m^2)
$A_{\mathrm{s}}$	shell flow area (m^2)
$c_{\mathrm{p}}$	specific heat capacity (J/kg K)
$C_{\mathrm{L}}$	laminar heat transfer geometry function (−)
$C_{\mathrm{T}}$	turbulent heat transfer geometry function (−)
CSP	concentrated solar power (−)
$D_{\mathrm{h}}$	characteristic diameter for Nu and Reh (m)
$D_{\mathrm{P}}$	characteristic diameter for ReP (m)
$D_{\mathrm{S}}$	shell inner diameter (m)
$D_{\mathrm{T}}$	tube outer diameter (m)
$f_{\mathrm{F}}$	fanning friction factor (−)
$\mathit{EPC}$	engineering procurement & construction
$h$	shell-side film transfer coefficient (W/m^2K)
$\mathit{HTF}$	heat transfer fluid
$K_{\mathrm{b}}$	hydraulic loss coefficient of baffle (−)
$L_{\mathrm{T}}$	length of tubes (m)
$N_{\mathrm{B}}$	number of baffles (−)
$N_{\mathrm{T}}$	number of tubes (−)
$\mathrm{Nu}$	Nusselt number (= $\frac{h}{k}{D}_h)$ s(−)
$Pr$	Prandtl number (= $\frac{c_p{\mu }_b}{k})$ (−)
$P_{\mathrm{T}}$	tube pitch (m)
${Re}_{\mathrm{h}}$	heat transfer Reynolds number (−)
${Re}_{\mathrm{p}}$	parallel flow Reynolds number (−)
$U_{\mathrm{B}}$	baffle velocity calculated from Ab (m/s)
$U_{\mathrm{s}}$	shell-side velocity calculated form As (m/s)
$∆P$	pressure drop (Pa)
$∆P_B$	baffle flow pressure drop (Pa)
$∆P_{\mathrm{L}}$	longitudinal flow pressure drop (Pa)
$k$	shell-side fluid thermal conductivity (W/m K)
${\mu }_{\mathrm{b}}$	bulk viscosity (Pa s)
${\mu }_{\mathrm{w}}$	wall viscosity (Pa s)
$\rho$	mass density (kg/m^3)