Mixing Util Help
Mixing Util Help
transfer utilities
Latest update: 29 September 2017
support@scale-up.com
This document summarizes the scope and features of DynoChem vessel utilities, including the ability to
incorporate an Excel vessel database of your organization’s equipment and to estimate the effects of impeller
and baffling configuration on mixing and heat transfer.
Estimated maximum power per unit mass in the impeller region / trailing vortex ..............................................9
Effects of baffles/ internals on liquid surface shape and power input ................................................................... 11
Applicability of correlations..................................................................................................................................... 13
Scaling rules......................................................................................................................................................... 14
Design space........................................................................................................................................................ 17
Applying utilities for mixing and heat transfer assessment and design .................................................................. 18
Usage if your organization has not set up its vessel database yet, or if your vessel is not listed in your database
............................................................................................................................................................................. 19
Porting your existing VesselDB worksheet into the updated vessel database file using our template ............. 20
References ............................................................................................................................................................... 26
Appendices .............................................................................................................................................................. 27
Appendix 3: Calculation of the effects of baffling on the liquid surface shape .................................................. 29
Significant progress has been made towards understanding and predicting fluid mixing in the period since the
1960s and much of this has been summarized well in reference 2, a textbook that we recommend to customers.
Several of the Scale-up Systems team conducted their PhD studies and industrial research on fluid mixing and
we continue to monitor and encourage publication of mixing research to enhance the knowledge base.
When we see opportunities to enhance or improve accuracy in the utilities, we do so by making changes and
sharing the new files in DynoChem Resources. We increment the version numbers and log the changes both in
the WhatsNew document on the website and also in the version history inside each tool. This means that
calculated results for a specific reactor system may change a little between older and newer utilities. We
recommend that users always download the latest utilities in order to benefit from these changes.
The following utilities are available and read vessel geometry and related data from an Excel-based customisable
vessel database contained within each utility:
Name of utility Utility estimates the following main parameters from vessel geometry and operating
conditions
Estimate_UA_DB.xls Heat transfer coefficient (U), wetted area (A), addition time, heating/cooling rate
Solid-liquid_DB.xls Power input (P), power per unit mass ( minimum agitator speed for suspension (NJS), in
solid-liquid systems, shear rate, e.g. solid-liquid reactions, crystallization,
hetereogeneously catalysed reactions
Single-phase_DB.xls Power input (P), power per unit mass ( mixing time (t95) – relevant for fast additions,
fed batch mixing/ mesomixing/ micromixing times – relevant for slow additions, shear
rate in single-phase/ homogeneous/ miscible liquid systems
Drawdown_DB.xls Power input (P), power per unit mass ( minimum speed for gassing (Ngas), mass transfer
coefficient (kLac) in vortexing and self-aspirating (hollow agitator shaft) gas-systems,
especially hydrogenation with headspace addition of H2
Gas-liquid_DB.xls Power input (P), power per unit mass ( mass transfer coefficient (kLac) in sparged (with
submerged gas addition below bottom impeller, though ring sparger) systems, especially
tall tanks with multiple impellers, e.g. fermenters
Liquid-liquid_DB.xls Power input (P), power per unit mass ( minimum speed for dispersion (NJD) in two-
phase/ immiscible aqueous/organic systems, e.g. extractions, liquid-liquid reactions
1. A new Design space feature has been included in several utilities that calculates process results over a
user-defined range of impeller speed and liquid volume. When applied to both Vessel 1 and Vessel 2,
this allows identification of a range of operating conditions in each vessel that lead to similar calculated
mixing parameters. Design space buttons are available on the Results worksheets and produce
response surface plots.
2. We have enhanced Vessel 1 and Vessel 2 Reports, including the user’s name, the date and the version
number of the utility. Reports now also contain individual impeller power numbers, UA intercept and
UA(v) where applicable.
3. We have extended our standard list of impellers, including the two-bladed flat paddle and a marine
propeller.
4. Users can now name, include and define multiple custom/user-defined impellers on the Impeller
properties tab; vessel database custodians can define a custom impeller list for use across an
organization. Custom types of baffles and internals can also be defined.
5. Users can easily import their organization’s vessel database (including custom impeller, baffle and
internals types) from a file on the network, Intranet or web site. This means that all users can apply the
latest utilities from DynoChem Resources and there is no need for power users / custodians to make
separate copies of the utilities and share them for internal use.
6. Unbaffled Power number estimates have been enhanced and made a function of Reynolds number.
7. We have added calculation of an estimate of the maximum power per unit mass generated by impellers
in a vessel, based on calculations related to the trailing vortex produced by the blades.
8. We have added calculation of torque per unit volume, a parameter sometimes used in systems with
higher viscosity systems and by agitator vendors.
9. We have added the Grenville, Mak and Brown (GMB) correlation (reference 13) as an alternative to
Zwietering for solids suspension with axial and mixed flow impellers.
10. Some worksheets are partially protected to prevent unintended or inappropriate edits by users,
especially the results calculated and displayed in blue text on a grey background. There is no password
and protection can be removed using Review>Unprotect sheet from the Excel menu.
For further reading, there is a list of textbooks (on mixing and other topics) that we recommend to our user
community in this document.
Our utilities are focused on agitators mounted on a single shaft on the central axis of cylindrical vessels, with
baffles fixed to the vessel wall or mounted separately in the vessel. These are the most common configurations
used in chemical processing.
Vessels are also normally fitted with a heat transfer ‘jacket’, in the form of an annulus around the base and side
walls. One important feature of the jacket is its wetted area, denoted A.
The configuration in Figure 1 provides a robust set-up for mixing of liquid mixtures with low viscosity, .
Impeller / agitator rotational speed is the main driver for mixing and this is represented by the symbol N. Usage
of four flat baffles fixed to the vessel wall helps to minimize rotation of liquid and vortex formation and makes
In the pharmaceutical industry, the set-up in Figure 1 is not used frequently. There are many reasons for this,
including:
In pharmaceutical applications, it is common to have just one or two baffles, only one impeller submerged and a
relatively low liquid level. These factors can make mixing less effective and also less predictable than in
configurations like Figure 1. The further you deviate from Figure 1, the less predictable your system is likely to
be.
Turbulent flow
To determine the flow regime in a mixing vessel, it is common to use the fluid mechanics concept of ‘Reynolds
Number’, a ratio of inertial to viscous forces. For an impeller mixing a fluid with density :
𝜌𝑁𝐷 2
𝑅𝑒 =
𝜇
DynoChem utilities are designed for working with low viscosity, turbulent flow systems stirred by relatively small
(also called ‘turbine’) impellers running at relatively high rotational speed, N.
Some parts of some utilities may be applied to systems with Reynolds numbers (Re) that are too low for full
turbulence (e.g. inside jackets, or for some calculations in small vessels). In general, results from DynoChem
mixing utilities are not directly applicable to non-turbulent flow in vessels or with non-Newtonian fluids; for the
latter systems (e.g. pastes, polymers) the viscosity depends on shear rate and may be quite high near the walls
even it if is quite low near the impeller. Positive-displacement impellers (such as a large anchor, helical ribbon
and various ‘gate’ designs) are often selected when dealing this these materials.
Power input
An important performance characteristic of an impeller, Po or ‘power number’, reflects the amount of power
the impeller transmits to the fluid. This is a type of ‘drag coefficient’, analogous to that of an airplane wing, and
may be used to calculate the power input (e.g. in Watts):
𝑃 = 𝑃𝑜𝜌𝑁 3 𝐷 5
6370
𝑅𝑒𝑐𝑟𝑖𝑡 = 1
𝑃𝑜 3
DynoChem utilities provide Po estimates for turbulent systems only. Po depends on impeller type, design, size
and position in the vessel as well as the degree of baffling. Under turbulent flow conditions, Po in a fully baffled
vessel becomes constant and independent of Re. Our correlations for fully baffled Po are described in Appendix
4.
The graph below (Figure 10.5 from reference 7) illustrates the dependence of Po on Re for a range of impeller
types.
𝑃
𝜀=
𝜌𝑉
Within a mixing vessel, the rate of dissipation of turbulent energy is highest near the impeller blades and lower
in the bulk of the vessel and especially near the liquid surface. It is possible for this parameter to vary by as
much as a factor of 100 between these high and low regions. The high region is therefore usually very small and
local can easily be 10 or more times the average in the vessel. The trailing vortex produced by the impeller is
considered the likely location of this maximum value and DynoChem utilities estimate this maximum value as
described below.
When conducting chemical reactions that occur very quickly as the fluid enters the vessel, addition of reagents
into the impeller region can be beneficial, so that fluid quickly experiences something like this high level of
energy dissipation. When dispersing liquids, the drop size produced can depend on this maximum energy zone,
especially if the droplet phase has a low concentration or the droplets are resistant to coalescence. Users
should be careful in general not to produce droplets that are very small, as these may take a very long time to
separate, especially at large scale, once the liquid-liquid process is complete.
Estimated maximum power per unit mass in the impeller region / trailing vortex
Recently, it has become possible to make an estimate of the local maximum energy dissipation rate in the
impeller region and especially the flow associated with the trailing vortex produced by most turbine impellers.
Measurements of the maximum levels of turbulent kinetic energy together, empirical correlations and
assumptions about the size of the energy containing eddies allow the following formula to be obtained (ref 11):
where x is a parameter representing the relative size of the trailing vortex and the impeller blade width. Values
between 12 and 17 have been reported and in DynoChem utilities we use x=15 for all impeller types.
Shear rate
Shear rates (spatial velocity gradients) vary a lot in a mixing vessel and the situation is further complicated by
the existence of both turbulent (fluctuating) and laminar shear. For your convenience, we provide the link to a
web preview of a section of reference 2 that deals with shear rates here.
In the utilities, the equations we use for shear are listed on the Basis tab (see below) and in the Excel comments
on Results tab cells.
The other estimates use power per unit mass and viscosity; they differ in the amount of power and the mass of
fluid considered. This leads to vessel averaged (all power in full liquid mass) and impeller averaged (30% of
power in impeller swept volume) turbulent shear rates. When users are concerned about particle breakage for
example, they sometimes review these turbulent shear estimates. However, it would be an oversimplification to
say that these shear estimates relate directly to breakage rates; other factors like circulation time (how often
particles pass through the impeller region) would also be relevant.
Reference 9 gives an impression of the range of equations in use to estimate shear rates. Reference 10 indicates
that about 30% of the impeller power is dissipated in the impeller region.
𝑄 = 𝐹𝑙𝑁𝐷 3
Under turbulent flow conditions, this parameter has been found by experiment to be related to Po (references 5,
6). For axial flow impellers:
1
𝐹𝑙 ≈ 0.7𝑃𝑜 3
Flow numbers in DynoChem utilities are based on these correlations for each impeller type, calculated under
fully baffled conditions and operating in the turbulent flow regime.
𝑀 𝑃
𝜏= =
𝑉 2𝜋𝑁𝑉
Calculations of torque per unit volume are included in several DynoChem utilities. This parameter is often
considered to be relevant when scaling mixing in systems with higher viscosity, which may have slow-moving
regions. It is also the basis for scale-up for maintaining fluid motion in liquids that exhibit a ‘yield stress’, i.e. are
semi-solid.
Top-to-bottom circulation is helpful when working with additions at e.g. the liquid surface that need to travel
quickly around the vessel. It is also helpful when additions should enter the impeller region directly, which is
often desired. Finally, top-to-bottom motion helps to overcome formation of layers that occurs due to density
differences between phases, e.g. in solid-liquid and liquid-liquid systems.
While baffling appears to help with distribution phases around the vessel, it does not always improve ability to
‘just suspend’ solids from the bottom of the vessel; this capability appears to be somewhat independent of
baffling, or at least no clear trend is evident (e.g. reference 14 which states “In general, the partial baffling
configuration appears to be the most successful in terms of achievement of the suspension state at the lowest
value of the agitation speed with the retreat curve-blade impeller”). On the other hand, there is no doubt that
baffling improves general mixing in a vessel containing a solid-liquid slurry, helping to mix the liquid as well as
distribute the solid particles.
Baffling is not universally helpful; for example, when solid particles are lighter than the liquid, formation of a
partial surface vortex with reduced baffling may be helpful; also when attempting to incorporate gas from the
headspace. In both of these cases, submergence of the top impeller is also an important factor and vessels may
only operate well over a narrow range of fill levels.
The effect of baffling is included in our utilities on the basis of baffling effect information from references 1 and
2 and interpolation of these results where necessary. We calculate ‘% baffling’ denoted p that depends on how
many baffling internals are submerged and the baffling effect of each. For more details, see Appendix 1.
Power numbers for unbaffled vessels in the turbulent flow regime depend on Reynolds number and we use data
from reference 12 to calculate unbaffled power number for each impeller, see Appendix 2.
The minimum baffling effect is 0% and the maximum is 100%. When there is 0% baffling, the liquid surface
shape is calculated using correlations for unbaffled vessels from reference 3, as described in Appendix 3. The
power number of each submerged impeller is then taken to be that of an unbaffled vessel and this configuration
produces the deepest vortex.
When there is 100% baffling, the liquid surface shape is assumed to be flat and the power number of each
submerged impeller is that of a fully baffled vessel. In this case, there is no vortex.
For intermediate levels of baffling between 0% and 100%, the liquid surface shape and power number of each
submerged impeller are calculated by interpolation between the results for unbaffled and fully baffled
conditions, using the percentage baffling as the interpolation parameter.
Further details are given in Appendices 1-4. Correlation parameters for all impeller types and the baffling effect
of each baffle and internals type are defined on the Impeller properties tab in each utility.
Figure 5 below from reference 1 and the caption support this assertion. However, in multi-phase situations (like
solid-liquid, illustrated), you can expect some asymmetry in how the phases are distributed.
You can therefore apply our utilities to get a rough estimate of performance in these configurations, as long as
you ‘model’ the system as partially baffled. A first approximation would be to model this using the normal
centred impeller set-up, with 1 baffle.
Laboratory reactors use some types of impeller that are rarely found in larger vessels, such as paddles with two
blades, half-moon impellers and magnetic stir bars. To apply the utilities to these systems will involve some
approximations and requires use of a user defined impeller row on the Impeller properties tab and input of
parameters to approximate these impellers.
Good correlation parameters for power input are quite difficult to find when the lab configuration is as small as
an EasyMax (e.g. 100 mL) and paddles may be non-standard compared to larger scale impellers. If you have the
possibility of measuring torque on the agitator shaft in such lab reactors (e.g. EasyMax or OptiMax), these would
be valuable data for the user community, so that Power number estimates can be made more accurate.
Each utility contains further calculations that are specific to that utility, to estimate mixing performance in the
context of a process result. For example, the single-phase utility calculates mixing time, the time taken to reach
homogeneity after a fast liquid addition; the solid-liquid utility calculates NJS, the minimum agitator speed
required to suspend solid particles. Correlations used for these additional performance calculations are
described inside the relevant utility (on the Basis tab and using Excel comments) and in a specific KB article listed
in DynoChem Resources containing further background information about the relevant type of mixing.
The solid-liquid, liquid-liquid and sparged gas-liquid utilities calculating mixing performance results (e.g. NJS, NJD,
kLa) that rely on engineering correlations applicable at larger scales; performance calculations are therefore not
expected to be accurate for small vessels (e.g. <10L). Some of the utilities include warnings of this type,
triggered by calculations of vessel volume.
The single-phase mixing, heat transfer / UA and gas drawdown utility should each give quite good estimates of
mixing performance (time, heat transfer, mass transfer coefficient kLa), even in small vessels.
For further reading, there is a list of textbooks (on mixing and other topics) that we recommend to our user
community in this document.
We provide a scaling rules dropdown list in several utilities that calculates in a separate column to the right of
Vessel 2, the agitator speed in Vessel 2 that would obey that scaling rule when compared to Vessel 1. To use
this feature, you should define the set-up and operating conditions of both vessels 1 and 2 on the Geometry and
Results sheets as fully as you can.
scale factor
1000
We recommend that users scale-up based on achievement of the required minimum agitation criterion that
affects the process result; such as ‘minimum speed to suspend solids’ in the solid-liquid mixing (NJS) utility. We
recommend that users also check that power per unit mass is in a similar range when scaling from vessel 1 to
vessel 2.
We provide other options in the dropdown list so that users can see how those would change the operating
conditions and mixing performance in vessel 2. Many users like to know what operating conditions would lead
to constant power per unit mass at different scales. Agitator vendors sometimes discuss or recommend scaling
at constant impeller tip speed. A shear rate scale-up might be used in a system where you were concerned with
damage to or breakage of materials in the process.
In all cases, we recommend reviewing power input per unit mass when scaling up and ensuring that this
parameter is in the range between 0.01 and 10 W/kg.
Axial flow
Radial flow
Mixed flow
The Impeller properties tab in the utilities lists the impeller types that you can select on the Geometry tab and
parameters in the equations used to estimate their performance characteristics. The general order of impellers
in the default/ standard lists is: radial flow impellers>mixed flow impellers>axial flow impellers>proprietary
impellers.
For each impeller, we calculate the power number under fully turbulent conditions as a function of the impeller
clearance and diameter. Further details are given in Appendix 4.
We include a standard impeller list covering turbines that are frequently used and covering the range from radial
flow (high Po) to axial flow (low Po). In general, other specific types of impeller that are not in our standard list
For example, marine propellers (with pitch=diameter) have similar solids suspension performance to hydrofoils,
though with a higher power number, e.g. 0.4 versus 0.3. Hydrofoil impellers were developed by agitator
vendors starting from the marine propeller design and looking to increase flow efficiency while also being lighter
and easier to fabricate.
Companies with many vessels that contain impeller types not on our standard list are encouraged to add their
custom impellers to a modified Impeller properties tab that their users can then import easily into our utilities
to model those systems. More details about the mechanics of doing this are given in the Vessel database
sections below.
In the drawdown utility, performance of a self-aspirating impeller may be estimated. This impeller has holes in
the blades that are connected to a hollow agitator shaft with a gas inlet in the reactor headspace. Rotation of
the blades creates a low pressure region behind the blade that induces gas from the headspace into the impeller
discharge region. In principle, self-aspirating impellers can be of any ‘type’; users should select the impeller
types that most closely resembles their self-aspirating impeller.
Each utility also includes an anchor impeller, due to the high frequency of occurrence of these impellers in some
regions. The anchor parameters we use are taken from reference 16 and relate to impellers mounted near the
vessel base with diameter equal to half the tank diameter or more. Anchor impellers are not very effective for
mixing in low viscosity systems and are likely to result in mostly tangential flow, with little top-to-bottom motion.
It is possible to include a Tickler (or ‘kicker’) impeller as the bottom impeller in all utilities. If present, this
impeller is usually very small and mounted at a low clearance. Its function is to keep the system (e.g. slurry)
agitated a little while the vessel is draining, so that all material can exit via the bottom valve. A Tickler impeller
does not make a significant contribution to mixing in the vessel under normal conditions.
When you include a Tickler as the bottom impeller on your Geometry tab, we include its contribution to the
total power input to the system, together with any other impellers.
The Tickler impeller type receives special treatment in the solid-liquid mixing utility only. For solids suspension
when a Tickler is present, we calculate performance (including S or Z in the Zwietering or Grenville, Mak and
Brown correlations, respectively) based on the second (upper) impeller if one is present and submerged.
Otherwise we use S=100 for the Tickler impeller, to highlight that this is very unlikely to help suspend solids
under normal vessel operating conditions.
If you would like a report that contains more detail than the report tabs, one option is to select the geometry
and results sheets and then from the Excel menu print the sheets. Print areas are set up in the utilities so that
this works easily and you can fit each sheet to a single page (portrait or landscape) to generate a PDF or
hardcopy for single set of input conditions in each vessel.
If you are using the utilities as part of a team project or technology/process transfer to another location, you will
find it useful to share the entire utility with your colleagues, e.g. by email attachment, on a file share or through
your document management system.
Design space
Several utilities include a capability to vary impeller rotational speed and liquid volume in vessel 1 or vessel 2
and generate a table of results for a selected mixing performance criterion, like power per unit mass. This
capability runs 17x17 simulations over a user-defined range of volumes and speeds, with an equal spacing
between conditions for each run. The table of results appears on the corresponding Vessel Rep tab (1 or 2)
together with a contour plot showing the response surface.
This capability may be used to define acceptable ranges of rotational speed and volume for a specific process, or
to identify the range of conditions in one vessel that match a certain response in another vessel.
Design space calculations should take less than a minute to complete and if you find that they are taking longer,
close any workbooks that do not need to be open.
Tip: We recommend that when saving the vessel utilities on your own computer, you save them as .XLs files, i.e.
in Microsoft Excel (97-2003) format, to preserve full functionality on reopening later. If this does not work as
you expect, try .xlsm (macro enabled) format as an alternative.
Tip: Working in Excel 2007 or later, you may occasionally notice that graphs in a utility appear to have been
moved or resized. To correct this Excel bug, use the zoom control to zoom e.g. in and then back to the previous
zoom level.
If your organization has already assembled its vessel database (containing VesselDB and Impeller properties
worksheets - this is highly recommended) you can import this database as follows:
1. Enter the location (full network address of the file, including its file extension, or a web address) of the
vessel database as shown below [get this from your vessel database custodian]
2. Press the Import button
A message will appear to confirm the import. You can then select your desired vessel in the dropdown list and
proceed with your design calculations.
When you select a vessel, relevant data are read from the database onto the Geometry and Results tabs.
If you override the database values, the blue text will change to black,
indicating that you have changed the settings relative to the database. This
also breaks the link between the field you have changed and the database.
To restore all settings in the utility to those in the database and restore the links to the
Restore
database, click the Restore buttons provided on the Geometry and Results tabs.
Usage if your organization has not set up its vessel database yet, or if your vessel is
not listed in your database
If your organization does not yet possess a vessel database, or the vessel that you need to model is not listed in
your vessel database, you can still use the utilities immediately as follows:
1. Select one of the pre-loaded vessels from the dropdown list, or select the
blank template for a user-defined vessel
2. Edit this vessel on the Geometry tab so that it represents your vessel
Tip: If you set up a user-defined vessel and want this to be added to the vessel database by your custodian/
administrator, send a copy of the utility to the custodian with your completed geometry and Results tabs.
You should create your database using our database template, downloadable here. If you already have a vessel
database that predates 2017, you will need to update your database in order for it to be used with our 2017
utilities, as described below.
Porting your existing VesselDB worksheet into the updated vessel database file
using our template
If you already have an existing vessel database, you should align the format of its VesselDB worksheet with our
standard format, which has not changed since 2013.
The VesselDB sheet has the same basic structure as in prior versions, with no added columns after DB (same
number of columns since November 2013). In the instructions below and in the database import tools
contained in the 2017 utilities, we assume that your existing database follows this same structure. If not, you
will need to edit and/or move columns in your existing database so that they match our template. Extra
columns after DB are not read or processed within DynoChem utilities. Column content from columns A to DB in
your database must be consistent with our template.
We recommend that you do not insert or delete rows in the VesselDB worksheet, as this may eventually impact
correct behaviour of Excel validation dropdown lists.
The ability to define multiple custom impeller types and use these in the droplists of your VesselDB sheet has
required minor changes in the validation lists on the VesselDB sheet:
Validation droplists for impeller type now point to named ranges that are defined on the Impeller
properties tab; this is necessary so that newly added user-defined impellers can be referred to
consistently and immediately on both worksheets (and where necessary inside our utilities).
Old validation droplists for impeller type located below row 600 on the VesselDB have been removed;
this part of the VesselDB sheet still contains needed droplists for other information such as units,
materials of construction and heat transfer fluids.
New in 2017, your vessel database workbook should contain an extra worksheet named Impeller properties, in
the same format as our template. This worksheet allows you to edit our default parameters for Impeller
properties and to add your own user-defined impeller types for use in your VesselDB sheet and our utilities, a
great new feature that many users have requested. The validation lists for impeller, baffle and internals types
on the VesselDB sheet in our template and each utility now point to ‘named ranges’ on the Impeller properties
tab, whose content updates dynamically when you add a user-defined impeller.
Open the utility or spreadsheet that contains your existing master vessel database.
o Check and if necessary fix any issues with column content consistency with our template, up to
and including column DB
o On the VesselDB tab, select rows with the vessel data that you would like to copy across. Be
sure to select entire rows using row headings at the left margin. This will also carry over any
supplementary information you have stored after column DB. Copy.
o Switch to our template. On the VesselDB sheet, select row 5. Paste Special… Values.
Repeat the above sequence for all of the vessels that you would like to import from your existing
VesselDB sheet.
As you have used our template to make this new vessel database, you automatically have an Impeller
properties tab in your file. This contains the standard impeller list and also space for you to add custom
/ user-defined impellers as described above.
Some customers have made the vessel titles in column B of the VesselDB worksheet into hyperlinks that
lead e.g. to vessel drawings on other worksheets in the same workbook. You can restore this type of
functionality in our new vessel database template by carefully copying the old vessel titles (with
hyperlinks) at the end of the above process, then selecting all of your ancillary worksheets containing
additional information and moving them to the new workbook. If your old workbook is .xlsx format, you
will need to save the new workbook as .xlsm, then close and reopen Excel before being able to move all
of the ancillary worksheets.
See the section below for advice on how best to deploy/ share this updated vessel database file.
You are welcome to send your completed vessel database to support@scale-up.com so that we can check its
format, after you have used our template to comply with the standard format.
When adding a vessel, if you would like to keep our initial list of vessels intact, type in the “blank” rows following
the initial examples. These “blank” rows have been populated with default values. As you add vessels, do not
leave additional blank rows between vessels or blank vessel names in the list because this may interfere with the
correct functioning of the Select Vessel dropdown menu on the Geometry sheet.
We recommend you list your vessels in a sequence that makes it easy for the users to find them when using the
Select Vessel dropdown menus on the Geometry tab.
Several of the cells use data validation to restrict the entered values to lists of valid choices, e.g. units of
measure, base types, impeller types, etc. You should not enter values other than the ones listed. Note that
using copy and paste into these cells will override the validation checking. You can find the complete collection
of validation lists by clicking on the link in cell I1.
We recommend that you do not insert or delete rows in the VesselDB worksheet, as this may eventually impact
correct behaviour of Excel validation dropdown lists.
If you are unable to provide values in the user defined section for some characteristics and these are important
for your work, consider approximating the characteristics for your impeller, based on values already in the list
for another impeller. You are welcome to contact support@scale-up.com to request our advice on parameter
values or assumptions that you make. We cannot guarantee that we will be able to recommend or help you to
refine your parameters.
One of the characteristics required is whether the impeller can be located at any location on the agitator shaft
or only as the bottom impeller. When you have defined this characteristic (in column BH), your new impeller
becomes part of the system and will (after importing to a utility) appear in the relevant dropdown lists on the
Geometry sheets of the utilities and in the dropdown lists on the VesselDB worksheet for defining your vessel
database.
The names of all user defined impellers must be unique in the impeller type list.
17 2 finger baffles 40
18 User defined 0
19 User defined 2 0
The names of all user defined baffle or internals types must be unique in the corresponding list.
Before the import proceeds, some error checking of the vessel database file is carried out:
It may seem attractive to advise users to simply copy and paste Impeller properties and VesselDB worksheets
from a vessel database into a utility but we advise against this approach, as this is generally prone to errors and
for example, validation lists for impeller type will not then point correctly at named ranges on the Impeller
properties tab.
The custodian should communicate the location of the vessel database file to potential users of the utilities.
This can be accomplished using the usual communication channels, such as email, or shared on a collaboration
website or wiki, if such a website exists within the organization and users visit there. The location should
include the full network/ file system pathname of the file or the full address of the web link / URL of the file, e.g.
The file indicate in the second image above is located at https://s3-eu-west-1.amazonaws.com/dc4/DB.xls. This
is the standard template also available from DynoChem Resources. The file cannot be imported to a utility
directly from the DynoChem Resources link because that link requires authentication (username, password).
The Amazon Web Services link does not require authentication.
Whether you use a file path or a URL, it is important that users do not require any additional authentication in
order to access the file location. The code behind the Import button does not include any authentication (such
as username or password).
Scale-up Systems also offers a service to provide a DynoChem TeamShare site for customers, hosted on
Microsoft Azure web infrastructure. This site is visible only to your organization and appears automatically for
users when they sign in to DynoChem Resources. This makes it easy for the custodian to distribute information
such as the location of the vessel database on your file system to all of your user community.
This is an additional task for the custodian that can quite easily be forgotten in busy times and can create
confusion for users about which utilities to use, the new ones in DynoChem Resources or older ones saved on
your own internal systems. As our utilities are enhanced regularly, we recommend always using the latest
versions for new calculations.
If you decide to adopt this more manual approach to managing your vessel database, the resulting custom
utilities can be saved and shared via the intranet, file share or a website. Once again a TeamShare site is a good
vehicle for this deployment, as it appears automatically when users log in to DynoChem Resources.
For general information on this approach, there is a helpful recorded webinar available in DynoChem Resources
entitled Incorporate your own equipment in DynoChem utilities.
[Additional sources of parameters are listed on the Impeller Information tab in each utility].
We denote the degree to which a vessel is baffled as ‘percentage baffling’ and this varies from 0% to 100%:
0% corresponds with no baffles or internals (other than the agitator and impellers) and allows rotation
of the liquid in the vessel, producing with a low impeller power number that also reduces with
increasing Reynolds number.
100% corresponds with four (4) flat baffles of width T/12, mounted T/60 from the vessel wall and at
ninety degrees to each other; this configuration is generally referred to as ‘fully baffled’ and at
moderate impeller submergence produces a flat liquid surface, reduces liquid rotation and the impeller
power number is at or near its maximum value. Power number in this set-up is independent of Reynolds
number once the flow is turbulent (typically Re>10,000).
The concept of percentage baffling is useful because it allows us to ‘bound’ impeller and liquid surface
behaviour between two well defined extremes and to interpolate behaviour at intermediate levels of baffling
between these values. Degree of baffling (as a fraction) is calculated using the following equation and may also
be entered directly by the user for calculations in the Vessel 1 or Vessel 2 columns on the Geometry tab of the
utilities:
𝑝 = 𝑚𝑎𝑥(∑𝑛𝑏
𝑖 𝛽𝑖 , 1.0) (1.1)
where i is the fractional baffling effectiveness of each baffling item and takes account of baffle type and
whether the baffle is submerged or not. Baffling effects of each item are therefore assumed to be additive, so
that the effect of adding more baffling items is captured. Baffling effects are defined on the Impeller properties
tab of the DynoChem mixing utilities, just below the section for impeller characteristics.
from which:
𝑅𝑃𝐷−0.33
𝛽= 1−0.33
(1.3)
Each item of baffling (e.g. flat baffles, beavertail, dip pipe) has a value of that is used in equation 1.3 to
calculate the overall baffling effect, p, of all submerged internals in a vessel from equation 1.1. That overall
baffling effect is used as defined in the sections below to interpolate liquid surface shape and power input
between the limits of zero and 100% baffling.
Importantly, the observation above about relatively high RPD with axial flow impellers and relatively low RPD
with radial flow impellers is reflected by this approach, as we calculate unbaffled power numbers for impellers
on an individual basis depending on impeller type and the lower relative values for radial flow impellers are
calculated automatically using the analysis below.
Unbaffled power number is calculated based on data presented in reference 12, which shows that:
The multiplier K and exponent n in equation 2.2 both vary with impeller type.
We have further analysed these data and expressions to obtain a set of equations for unbaffled power number
that apply to all types of turbine impeller studied in reference 12, from high power number radial to low power
number axial flow impellers:
𝑏_𝐾
𝐾𝑃𝑜 = 𝐴𝐾 𝑃𝑜𝑝=1.0 (2.3)
𝑏_𝑛
𝑛𝑃𝑜 = 𝐴𝑛 𝑃𝑜𝑝=1.0 (2.4)
Parameters K and n can therefore be estimated knowing only the power number under fully baffled conditions.
Figure A.2 below shows the fit of 2.3 and 2.4 to the data in reference 12.
From the regression shown in Figure A.2, AK=1.9823 and bK=0.845; An=0.1483 and bn=0.1523.
This additional analysis allows an estimate of the unbaffled power number to be made for an impeller based
solely on knowing its fully baffled power number. This is useful because pharma vessels may contain impellers
(including user defined impellers) that are less commonly used in other parts of the chemical industries and
therefore also less well characterized.
To summarize, we use the above parameter sets A and b to calculate K and n and hence to estimate Pop=0 for
each impeller, using Pop=1.0. We then apply equation 2.1 with p to obtain Pop for the current vessel operating
conditions.
If p =degree of baffling (between 0 and 1), calculated from equation 1.1 above and the number and type of
baffles and their submergence, we calculate the partially baffled vortex depth (hv)p from the unbaffled vortex
depth (hv)u as follows:
Once the vortex depth is known, the vortex shape can be calculated assuming a Nagata vortex (combination of
free and forced vortex) with conservation of liquid volume as a constraint. This allows the vortex shape and
liquid height at the wall to be calculated.
If the vortex depth is located beneath the vessel, the intersection of the vortex and the vessel base needs to be
calculated, and the shape is calculated iteratively to conserve volume.
For illustration, results produced by the utility have been compared with photographs from reference 4 in the
table below.
N=300 rpm
N=522 rpm
Table 3.1: Vortex depth photographs and predictions: unbaffled vessel, 0.29m diameter, reference 4.
In order to calculate the vortex depth in partially baffled systems, the unbaffled vortex depth is first calculated:
𝑇 𝑏6
ℎ𝑣 𝑏2 𝑇 𝑏3 𝑏4 𝐺𝑎 𝑏5 ( )
𝑑
= 𝐵1 𝐺𝑎 (𝐷) 𝐹𝑟 𝐷 (3.2)
Coefficients B1,b2,b3,b4,b5,b6 are obtained from reference 3. In equation (3.2) the symbols mean:
D=impeller diameter
T=tank diameter
hv=vortex depth
User defined
Retreat curve impeller* 1e8 2e10 1.23 0.069 -0.38 1.14 -0.008 0.008
Curved blade turbine* 1e8 2e10 1.23 0.069 -0.38 1.14 -0.008 0.008
3-bladed pitched blade turbine 1e8 2e10 0.84 0.069 -0.38 1.14 -0.008 0.008
4-bladed flat blade turbine (est) 1.32 0.069 -0.38 1.14 -0.008 0.008
6-bladed flat blade turbine 3e7 5e10 1.52 0.069 -0.38 1.14 -0.008 0.008
Rushton disc turbine 3e7 5e10 1.52 0.069 -0.38 1.14 -0.008 0.008
Intermig
6-bladed pitched blade turbine pumping down 8e6 1e10 1.13 0.069 -0.38 1.14 -0.008 0.008
Chemineer CD-6* 3e7 5e10 1.52 0.069 -0.38 1.14 -0.008 0.008
*Hydrofoil, Turbofoil, Maxflo Y and HE3 assumed the same as 3 blade PBT; CD6 and BT6 assumed same as Rushton disk
turbine; RCI and Curved blade turbine assumed same as 3 bladed flat blade turbine estimated (B1 is 0.29 less than a 6 blade
flat blade based on difference for 6 and 3 blade PBT). No values available for an Intermig or an up-pumping impeller (no
vortex formed).
Power number correlation parameters fitted by Scale-up Systems to measured published impeller data for fully
baffled conditions are stored on the standard Impeller properties tab in each utility and the standard vessel
database template, together with all other impeller performance calculation parameters.
Solids suspension correlation parameters fitted by Scale-up Systems to measured published impeller data are
also stored on the Impeller properties tab, for calculation of the ‘s’ parameter in the Zwietering equation. The ‘z’
parameter for use in the Grenville Mak and Brown solids suspension correlation is also stored there for axial and
mixed flow impellers.
We would welcome submissions of further Po or s data or references that we can use to improve these
correlations. Both the fully baffled power number and solids suspension calculations currently use a linear form
of equation to calculate the effects of impeller diameter clearance and degree of baffling (p, where applicable)
on Po and s:
𝐷 𝐶
𝑃𝑜𝑏 = 𝑃𝑜𝑖𝑛𝑡𝑒𝑟𝑐𝑒𝑝𝑡 + 𝐴𝑃𝑜 𝑇 + 𝐵𝑃𝑜 𝑇 (4.1)
𝐷 𝐶
𝑠 = 𝑠𝑖𝑛𝑡𝑒𝑟𝑐𝑒𝑝𝑡 + 𝐴𝑠 𝑇 + 𝐵𝑠 𝑇 +𝐶𝑠 𝑝 (4.2)
This approach has been found to fit satisfactorily to a range of impeller data; 95% confidence intervals are within
20% of the predicted values. Using a logarithmic function instead with exponents on the parameters did not
improve the fit. No empirical correlation will work perfectly over a significant range of these parameters; given
the variability within the measured data, this approach appears to be reasonably accurate.
When the user’s impeller configuration is outside the range of D and C covered by the correlations (4.1) and (4.2)
above, the indicator ‘E’ appears beside the calculated value on the Results tab. This indicates ‘extrapolation’
outside the region where we have measured data and the values used will be those that apply at the edge of the
covered range. When the configuration lies within the range of the correlation, the indicator ‘I’ for interpolation
appears.
For less common impellers, where fewer published data are available, we approximate their characteristics
based on similar impellers that are better characterized. We apply the correlations in (4.1) and (4.2) for these
less common impellers at conditions C/T=D/T=0.33. The ‘E’ indicator appears on the Results tab to indicate
extrapolation.
For multiple impellers, we treat the vertical distance between each impeller as a ‘clearance’ and apply equations
4.1 and 4.2.