HTRI’s continuing research in process heat transfer has produced changes to several methods. Summarized here are those that may affect the cases users will run in Version 9.1.

Recommendations for alternative boiling correlations to those based on reduced pressure

HTRI originally developed simple boiling correlations as a function of reduced pressure. Because the definition for true critical pressure is often uncertain for mixtures and process simulators, users often relied on a pseudo critical pressure definition which led to inaccuracies in these simple boiling methods. However, industry is increasingly using process simulators to provide detailed fluid properties, enabling HTRI to develop a more complex boiling model that handles more types of fluids over a larger operating range.

In Version 9.1, HTRI updated the following methods to an alternative boiling correlation, described in more detail in BG1-17

• Shellside and tubeside upper limit for nucleate boiling heat transfer coefficient
• Shellside critical heat flux
• Shellside film boiling methods
• Tubeside subcooled film boiling methods

The first change prevents overprediction of the nucleate boiling contribution at high reduced pressures, lowering the overall boiling coefficient for those cases. However, the other three changes result in an increased boiling coefficient. The overall impact on the boiling coefficient depends on which contribution is the dominant coefficient for that case.

Some HTRI boiling methods still rely on reduced pressure, but an option introduced in Version 9.0 uses the density ratio as a check. Users should consider using this option for cases where they provide the pseudo critical pressure. To activate this option when the true critical pressure is unknown, simply remove the critical pressure and/or specify the density ratio on the Control/Methods panel.

Additional reading: BG1-17

New Electric Heater service types

Xist^® 9.1 adds four new service types: electric heater, electric thermosiphon reboiler, electric forced flow reboiler, and electric kettle reboiler. These new types required three electric heater-specific inputs: heat flux, unheated length, and maximum sheath temperature. To run electric heater cases, Xist uses a constant heat flux boundary condition for shellside performance calculations.

The new electric heater capability is appropriate for general sizing, but additional research and software work is required to improve it. While its implementation in Xist automates the guidance from TechTip 10, it does not address issues such as the location of hot spots or the identification of a maximum sheath temperature for burnout.

Additional reading: Revised TechTip 10 (forthcoming)

Conversion of simple piping input option to detailed piping output results

Although users will still have access to the default simple piping input option for reboilers, the program now converts this input to a detailed reboiler piping geometry when the case is run. This removed the need for two separate code bases and clarified the assumptions that were used in the simple piping options based on the different input options. The conversion logic assigns appropriate inlet and outlet piping elements based on the specified TEMA type, geometry, orientation, and location of the boiling fluid, applying detailed piping methods to the converted inlet and outlet piping geometries, reporting the results as detailed piping outputs.

The conversion accounts for the inclusion or exclusion of the Bend allowance specification in the simple piping input. When Bend allowance in the simple piping specification is set to Yes, the program returns an actual pressure drop for the elements associated with the bend allowance. When Bend allowance is set to No, the program reports a pressure drop of 0 for the elements associated with the bend allowance.

This behavior continues if you select Enter detailed piping and click the Copy Detailed Piping Output to Input button. The program copies the converted detailed piping output results to inlet and outlet piping tabs for additional finetuning.

Detailed piping elements have also been improved:

• You can now specify the Eccentric reducer/expander length, and the program corrects the associated element pressure drop for the calculated angle. When no length is specified, the default angle is 18 degrees.
• The program prorates bend loss coefficients when 2000 < Re < 4000. Previously, a significant discontinuity occurred at Re = 2000, and the new method provides a smoother transition between laminar and turbulent flow.

Additional reading: Revised TT-19 (forthcoming)

Rate-controlled condensation of mixtures

The HTRI Resistance Proration Method (RPM) is the default mixture condensation method for plain and low-finned tubes in Xace^®, Xist, Xhpe^®, and Xjpe^®, as well as for the plate-frame configurations in Xphe^®. The RPM utilizes mixture correction functions to approximate a diffusion mass transfer process. When mass transfer by convection dominates, the condensation is “rate controlled”, and the RPM method in Versions 9.0 and earlier is conservative.

Recent research data collected under rate-controlled conditions indicate that the RPM underpredicts the heat transfer coefficient. These data differ from previous research data for mixtures because the amount of noncondensables is low and the mean temperature difference (MTD) is high. The mixture correction functions in Version 9.1 more accurately model the recent data. Adjustments have been applied to all RPM calculations both inside and outside tubes and in plate-frame configurations.

These changes should not significantly affect most industry cases. However, in selected cases with high mass transfer rates (such as when the MTD is high with multicomponent condensation), the condensing coefficient could increase more than 20%.

Additional reading: CS-23

Updates to shellside subcooled boiling methods in Xist

HTRI has updated the logic for how the program handles certain inputs when calculating the shellside subcooled boiling heat transfer coefficient.

• The program now respects the Nucleate only selection in the boiling methods panel and ignores contributions from convective boiling and natural convection. The program still prorates the contribution from the nucleate boiling coefficient appropriately when applied.
• The program replaces the contribution of the convective boiling coefficient with the sensible liquid coefficient, if specified.

Changes to reporting in the shellside monitor facilitate evaluation of the subcooled boiling coefficient.

• The shellside monitor now lists the natural convection coefficient for subcooled boiling increments.
• The reported sensible liquid coefficient is corrected for subcooled boiling increments.

Results are impacted only for shellside boiling cases when users specify either input, but the reported values now reflect the correct result in all cases.

New acoustic vibration check for economizers

Xace 9.1 checks for the occurrence of acoustic vibration in economizers using the HTRI-Chen method, similar to the approach Xist uses for shell-and-tube exchangers. The check evaluates two issues:

• Does the dominant frequency of the pressure fluctuations in the crossflow wakes match the natural frequency of the rectangular duct?
• Does the acoustic excitation have sufficient energy to form a standing wave across the duct?

Xist uses the Fitz-High Strouhal numbers and the Owens turbulent buffeting frequency to check for lock on to the fundamental acoustic natural frequency—a function of the speed of sound in the medium and the shell diameter. In contrast, Xace 9.1 computes an excitation frequency from Strouhal number relationships developed specifically from and for acoustic vibration in rectangular crossflow bundles. In the economizer, the length scale for a potential standing wave is the duct width across the bundle instead of the shell diameter. In fact, much of the research on acoustic vibration is performed in wind tunnels where the tube bundles under test are placed in rectangular ducts, similar to tube bundles in economizers.

Additional reading: Design Manual updates (forthcoming)

Updated shellside sensible liquid and liquid boiling heat transfer methods for non-baffled exchangers using plain tubes

Version 9.0 and earlier applies non-baffled shellside methods when you select None for the baffle type for E, F, G, H, and J shells. The default method prorated longitudinal and window flow models, and you could also select Pure longitudinal flow as an option on the Control/Methods panel.

HTRI has developed new shellside sensible liquid and boiling heat transfer methods for non-baffled exchangers with plain tubes, as described in BG1-15 (vertical tubes) and BG1-16 (horizontal tubes).Version 9.1 includes a number of changes:

• A longitudinal flow model replaces both the current default model and the pure longitudinal flow model.
• The window flow model is no longer used.
• Static pressure drop is now calculated for boiling cases with a vertical shell.
• A stratified flow model is implemented for horizontal boiling cases.

It is difficult to generalize whether the new methods increase or decrease the calculated heat transfer coefficient for any particular case. However, the previous default method overpredicted and the pure longitudinal flow underpredicted the coefficient in single-phase cases. We expect the stratified flow model to reduce the shellside coefficient for H-shell reboilers, generally by less than 10%. However, for wide boiling range mixtures, you may observe larger reductions.

Additional reading: BG1-15 and BG1-16

Xace natural draft method and mode consistency

In previous versions of Xace, the airside heat transfer coefficient depended solely on the Reynolds number. In Version 9.1, the airside heat transfer coefficient depends on the Rayleigh and Reynolds numbers. This mixed convection formulation better predicts heat transfer around high-finned tubes at lower airside flows. In general, the mixed convection method reasonably predicts heat transfer coefficients derived from HTRI data. Note: The method applies to horizontal tube bundle orientations with high fins.

Below are some other improvements in Version 9.1:

• A consistent correlation for both Air-cooled heat exchanger and Natural draft air-cooler modes combines forced and free convection.
• Natural draft air-cooler mode allows both Rating and Simulation calculations, whereas in previous versions, this mode allowed only Simulation calculations.
• Air-cooled heat exchanger and Natural draft air-cooler modes calculate
  • improved friction and contraction-expansion pressure losses
  • consistent pressure losses due to flow entrance, exit, and ground clearance
  • pressure losses due to momentum change between plenum-bundle sections in forced draft and induced draft
• Air-cooled heat exchanger and Natural draft air-cooler modes report tube bundle pressure drop and airside pressure loss details in a new Final Results Supplementary tab.
• Natural draft air-cooler mode now calculates
  • theoretical draft force
  • required chimney height

Additional reading: Design Manual updates (forthcoming)