HTRI's continuing research in process heat transfer has produced changes to several methods. Summarized here are those that may affect cases users run in Version 9.2.

New kettle circulation method and updated methods and modeling options

A new field on the Kettle input panel allows users to choose from Legacy, Overall, and 5 zone options. The Legacy method is the model present in previous versions of Xist^®. Selection of the Legacy model results in answers very close to previous versions. The Overall and 5-zone options represent a new circulation boiling method, described in BK1-17. The 5-zone option uses the same circulation boiling method as the overall option but converges on 5 zones, offering a better option for cases with a significant duty variation along the bundle.

The main changes in the circulation models are the following:

• The bundle height used for bundle pressure drop is now consistent with the height used to calculate the liquid static head driving force.
• A modified calculation of liquid volume fraction accounts for recirculating liquid inflow on both sides of the bundle, which depends on vapor generation.
• A calculated upper limit is imposed on the recirculation ratio.

In addition to the new circulation method, we also modified the heat transfer and pressure drop methods. The main changes are listed below:

• Changed from four-mechanism (nucleate, forced convection, natural convection, and thin film) to two-mechanism model (nucleate and convective), a simplification that provides a smoother balance between convection and nucleate boiling and links convective boiling calculations directly to recirculation
• Modified convective boiling methods to account for the inclusion of natural convection and thin film boiling
• Modified C factor used to calculate the two-phase frictional pressure drop in bundle
• Updated transition and film boiling methods, as documented in BK1-19
• Corrected bundle bubble point based on recirculation as discussed in BK1-16

With these changes, the methods better agree with HTRI experimental kettle data and produce a more realistic circulation ratio. The new circulation model significantly reduces the number of cases in which the legacy model predicts a recirculation ratio of less than 1. On average, we see an increase in the circulation ratio, resulting in a slightly increased heat transfer coefficient, except in cases of large kettles using low-finned tubes where the previously thermal dominating mechanism was thin film.

Impact of modifications

• Better agreement with HTRI experimental kettle data
• Significantly fewer cases predicting a recirculation ratio of 1 (e.g., no recirculation)
• Higher predicted recirculation ratio in general, especially at high vapor generation
• Better handling of cases with high subcooling at inlet
• Better predictions for small bundles

Additional reading: Reports BK1-15, BK1-16, BK1-17, and BK1-19; pending Design Manual changes

Updated heat transfer and pressure drop methods for falling film evaporation

Xist 9.2 includes revised tubeside falling film evaporator heat transfer and pressure drop methods based on additional experimental data taken by HTRI. Previously, Xist underpredicted the heat transfer coefficient for some tubeside falling film evaporator cases. In general, the improved methods calculate higher heat transfer coefficients and slightly higher pressure drops compared to those in previous software versions.

This version issues two new warning messages for tubeside falling film:

• Marangoni liquid film breakdown based on Reynolds number
• Maximum flooding flow rate at tube inlets based on Froude number

The criteria for issuing the existing warning messages about nucleate boiling film breakdown and mist flow dryout have also been revised to reflect improved methods.

To provide users more information for case evaluation and troubleshooting, the Tubeside Monitor for falling film cases includes additional items:

• Wall superheat required for onset of nucleate boiling
• Weber number
• Kapitza number
• Liquid film thickness

Additional reading: Reports BT-41, BT-42, and BT-43

Non-Newtonian flow heat transfer and pressure drop calculations included in Xist, Xphe^®, and Xjpe^®

Experiments were conducted to develop heat transfer and pressure drop predictions for non Newtonian fluids, and the results are documented in S-ST-1-16. In Xchanger Suite 9.2, Xist, Xphe and Xjpe have non-Newtonian heat transfer and pressure drop calculations for the tubeside liquid phase only. A power law model accepts multiple data points of temperature vs. consistency index and power law exponent, in the physical property input forms. Input fields are available only for tubeside liquid applications with no phase change. The new methods are valid for pseudo plastic fluids with no yield stress. For practical purposes, the validity of equations is limited to tubeside Reynolds numbers less than 1000. A CFD study and experimental data for multiple fluids showed good agreement with these methods. Modeling a non-Newtonian fluid with the new methods results in predictions of lower heat transfer and greater pressure drop, compared to Newtonian methods.

Additional reading: Report S-ST-1-16; Upcoming webinar: Using CFD Simulation to Assess HTRI Predictions for Non-Newtonian Fluids ; pending Design Manual updates

Updated pressure drop methods for tubeside downflow condensation

Two-phase friction and momentum pressure drop methods were updated for downward flow in tubes under deep vacuum. Models of condensation in these conditions are very sensitive to pressure drop from friction and momentum, with small changes in pressure resulting in large swings in saturation temperatures. The update in Xchanger Suite 9.2, based o9n research presented in CT-31, includes changes to the two-phase friction correction factor, transition from a separated flow model to a homogenous flow model under certain conditions, and changes to the momentum recovery factor.

The improved methods produce lower incremental pressure drops for two-phase downward flow under 14 kPa. Many cases that would previously fail due to the calculated pressure drop exceeding the inlet pressure now successfully produce a rating. For cases that would previously run, users should expect improved accuracy in the saturation temperature and more reliable models.

Additional reading: Report CT -31; pending Design Manual updates

Availability/Exergy analysis report

A second-law definition of efficiency can account for unavoidable sources of irreversibility in a heat exchanger—e.g., heat transfer over a finite temperature difference and flow friction—and is based on the destruction of flow availability, which is a function of the mass flow rate of the streams, the enthalpy (duty), the dead state temperature, and the entropy change. Currently, heat exchanger design approaches implement only conservation laws for mass and energy, i.e., the first law of thermodynamics. Moreover, the second-law analysis informs the environmental impact (i.e., the carbon footprint) of the operation of a process heat exchanger.

The new availability loss report contains results from a second-law analysis for the change in the availability of the exchanger. This report is limited to Xace^®, Xist, Xjpe, and Xhpe cases that use physical property methods that provide entropy. Users can specify the dead state used for this analysis on the Methods panel. For cases that previously ran, the report is produced only if fluid properties are generated again in Version 9.2.

Additional reading: Report RR-6 (pending publication); pending Design Manual updates; pending webinar

Revised definition of percent overdesign

In previous versions, % overdesign was calculated as 100 [(U_act / U_req) – 1], regardless of values calculated for U_act and U_req. A lower limit of -100% was applied if U_act was less than U_req. The interpretation of a negative percent overdesign became unclear as many users as well as convergence algorithms were interpreting it as the required area to consider.

Version 9.2 changed the definition of percent overdesign to the following:

• If U_act > U_req, % overdesign = 100 [ (U_act / U_req) – 1]
• If U_act < U_req, % overdesign = 100 [1 – (U_req / U_act)]

This change has no impact on positive overdesign cases.

Modified calculation of X-shell velocity

Xist divides X shells horizontally to produce increments for calculating conditions along the vertical flow path. When calculating the crossflow velocity for each increment, previous versions determined the crossflow area using the integral average chord length. The updated methods now use the local bundle chord length of each increment to calculate crossflow velocity, which allows a more appropriate velocity profile to be established.

For most cases, this change does not significantly impact the overall calculations, the incremental trends are more accurate as they progress from a smaller chord length to the middle of the shell. The cases that could be more significantly impacted are vacuum condensers with a significant amount of desuperheating. This updated method captures this higher velocity and therefore allows more accurate predictions of pressure drop and therefore mean temperature differences for these vacuum applications.

Additional reading: Pending Design Manual updates