MethodsX 11 (2023) 102410

Contents lists available at ScienceDirect

MethodsX
journal homepage: www.elsevier.com/locate/methodsx

Experimental determination of dynamic pseudo-equilibrium

moisture content: A practical limit for the drying process ✩
Alexis Manuel Faneite Noguera a,b, Ignacio Angós Iturgaiz c,∗
a
Laboratory of Chemical Engineering, Faculty of Engineering, School of Chemical Engineering, University of Zulia (LUZ), Guajira Avenue, Campus
“Dr. Antonio Borjas Romero”, Maracaibo 4001, Venezuela
b
Department of Research, Development & Innovation, Industrial Dryers of Venezuela Corporation (SECAVENCA), Maracaibo 4001, Venezuela
c
Public University of Navarre (UPNA), Department of Agricultural Engineering and Biosciences, Institute for Innovation & Sustainable Food Chain
Development (ISFOOD), Campus de Arrosadia, 31006 Pamplona, Spain

a r t i c l e i n f o a b s t r a c t

Method name: Simulation and rigorous design of industrial dryers combine a large number of models, which
Determination of dynamic pseudo-equilibrium feed three fundamental balances: (1) mass; (2) energy; and (3) quantity of movement of the ma-
moisture contents in an experimental drying terial through the dryer. Many of these models represent physical phenomena affecting the three
kinetics study, and calculation of the practical
balances at the same time, which makes these calculations extremely complex, hence, accurate
moisture ratio for the modeling of said
models are essential. The hypothesis that the kinetic stage of drying of any material culminates
experimental kinetics, for use in rigorous design
and simulation of industrial dryers in the thermodynamic moisture equilibrium between solid and drying gas has been in effect for
many years. However, recent findings show that there is a transition stage between the kinetic
Keywords: stage and the thermodynamic equilibrium, which, experimentally, looks like an equilibrium. The
Drying kinetics beginning of this transition stage or dynamic pseudo-equilibrium stage would mark the end of the
Moisture ratio drying kinetics models, which has been named as the dynamic pseudo-equilibrium moisture con-
Moisture fraction tents (Xdpe ). The non-observance of this phenomenon presupposes a model limited in its prediction
Equilibrium moisture content
capacity, especially in the last stages of drying and even more so at low drying temperatures. As a
Simulation of industrial dryers
consequence, sizes of industrial dryers could be underestimated during the simulation and rigor-
ous design process, or underestimate drying times, in batch dryers. On the other hand, the optimal
conditions may never be found, during the optimization of existing industrial drying processes.
The objective of this work is to present the procedure to determine Xdpe , during the experimental
determination of drying curves of any material. Likewise, to propose the practical moisture ratio,
which uses Xdpe , instead of the equilibrium moisture, to be used in the modeling of the drying
kinetics.

• The drying process is divided into three stages: kinetic, transition, and equilibrium.
• The dynamic pseudo-equilibrium moisture content divides the kinetic and the transition
stages.
• The practical moisture ratio should be used in rigorous industrial dryer design calculations.

✩
Related research article: A. Faneite, A. Rincón, A. Ferrer, I. Angós, G. Arguello, G., Mathematical modeling of thin-layer drying of green
plantain (Musa paradisiaca L.) peel, Int. Food Res. J. 23(5) (2016) 2088–2095. URL: http://www.ifrj.upm.edu.my/23%20(05)%202016/(33).pdf
∗
Corresponding author.
E-mail address: ignacio.angos@unavarra.es (I. Angós Iturgaiz).

https://doi.org/10.1016/j.mex.2023.102410
Received 27 February 2023; Accepted 29 September 2023
Available online 30 September 2023
2215-0161/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

Specification table

Subject area Chemical Engineering

More specific subject area Drying kinetics
Name of your method Determination of dynamic pseudo-equilibrium moisture contents in an experimental drying kinetics study, and calculation
of the practical moisture ratio for the modeling of said experimental kinetics, for use in rigorous design and simulation of
industrial dryers
Name and reference of original method Mathematical modeling of thin-layer drying of green plantain (Musa paradisiaca L.) peel
Resource availability No specific resources are required.

Method details

Background

Smith [1], in 1918, mentioned the term moisture ratio for the first time to refer to moisture on a dry basis (db), in units of kg/kg
(water/dry matter). However, Lewis [2], three years later, used it to refer to the moisture ratio (dimensionless). In his famous work,
where he published the first known thin-layer drying model, being the Lewis’ concept of moisture ratio, which remains valid to date.
The traditional theoretical moisture ratio (Mr), expresses the amount of water evaporated in time t, in relation to the total amount of
water that can be evaporated by drying, and whose obvious limit is the equilibrium (Eq. (1)).
𝑋𝑡 − 𝑋∞
𝑀𝑟 = (1)
𝑋0 − 𝑋∞
where Mr is de traditional theoretical moisture ratio, X is the moisture content on a dry basis (db) (kg/kg) (water/dry matter), and
the subscripts 0, t, and ∞ are the initial state, the state at time t, and the equilibrium state, respectively.
Mr was used to express the moisture content, in studies of drying kinetics, for 80 years. The drying kinetics data, in terms of
Mr, obtained from said studies of drying kinetics are correlated as a function of time to obtain the kinetic models. In 2001, Freire
et al. [3], had already required to correct Mr to consider an external resistance (initial period of thermal stabilization) that occurs,
at the beginning of the olive bagasse drying process. This modification to Mr calculations, however, was not generalized. In 2010,
Li et al. [4], found a pseudo-equilibrium period, baptized as practical equilibrium, which divided the dynamic period, and the real
equilibrium, in the osmotic dehydration processes of fruits. Since real equilibrium is extraordinarily hard to achieve, Li et al. [4],
instead recommend the use of practical equilibrium.
In 2016, Faneite et al. [5], found a region of the drying curves of green plantain (Musa paradisiaca L.) peel, at a moisture contents
below 10% on a wet basis, (wb), in which the initial monotonic decreasing slows down appreciably. In said region of the drying
curves, all the thin-layer drying models, lost their predictive capacity to a greater or lesser degree. The moisture content where
the change of the initial monotonic decrease as a function of time (or kinetic stage, which has a characteristic kinetic behavior) is
visualized, was named as dynamic pseudo-equilibrium (dpe) moisture content. In a subsequent study on thin-layer drying kinetics of
peeled cassava root slices (year 2020), Faneite et al. [6], also found the existence of this stage for this other material. More recently,
in 2022, Kappaun et al. [7], reported the drying kinetics of the green banana peel, including a part of the dynamic pseudo-equilibrium
stage, but erroneously, they classified it as the real equilibrium.
The dynamic pseudo-equilibrium moisture content is then a thermophysical property of drying, whose numerical value is normally
below the storage moisture content, and above of the real equilibrium moisture content. And the pseudo-dynamic equilibrium stage,
likewise, is a transition stage between the initial kinetic stage and the final stage of real thermodynamic equilibrium and can also
be called the kinetic/equilibrium transition stage. Its behavior, in terms of moisture content as a function of time, is approximately
an average between the mentioned stages. These findings imply that Eq. (1) can no longer be used, as an expression of the moisture
content, in the modeling of experimental drying kinetics. This is because the ultimate goal of modeling the thin-film drying kinetics
of any material of interest is to make the resulting model useful for rigorous design and simulation. As will be explained later, the
calculation of the output moisture of the solid from the dryer will be erroneously estimated by the simulator when it is below Xdpe .
Thus, for simulation and design purposes, Eq. (2) should be used [5,6], being the experimental determination methodology of
Xdpe , detailed in this method.
𝑋𝑡 − 𝑋𝑑𝑝𝑒
𝑝𝑀𝑟 = (2)
𝑋0 − 𝑋𝑑𝑝𝑒

where pMr is the practical moisture ratio (term proposed in this method), Xdpe is the moisture content (db) at the beginning of the
dynamic pseudo-equilibrium stage (kg/kg) (water/dry matter). Xt and X0 were already previously defined, in the definition of terms
of Eq. (1).
The prefix "pseudo" (false) comes from the fact that pseudo-equilibrium, seen in a small period of time, looks like a true equilibrium
state. However, when viewed in its entirety, it can be noticed that It is a transitory process of very slow moisture change over time.
This stage begins at the end of the typical kinetic process and ends when true thermodynamic equilibrium has been reached. It is
dynamic because it occurs in studies where there is air circulation, and not in traditional gravimetric studies of macroscopic static
equilibrium, where air does not flow over the material [8]. But in addition, it is dynamic because it preserves a small fraction of the

2
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

Fig. 1. Representation of the entire drying process, up to equilibrium moisture.

dynamic process of material transport, through the solid, typical of the initial kinetic stage [4]. The dpe moisture content is lower
than the storage moisture content (except at very low drying temperatures), but higher than the equilibrium moisture content.
The complete drying process, in terms of simplified moisture ratio (Eq. (3)) vs time (t) can then be visualized in Fig. 1, where the
process starts in a Xt /X0 of 1 at t0 . Next, a first stage of monotonic decrease in moisture content, widely described in the literature, is
established. Eventually, this curve reaches the dynamic pseudo-equilibrium simplified moisture ratio (Xdpe /X0 ), where the behavior
of the curve changes. At this point, the drying process enters in a transition stage between the kinetics and the equilibrium, which
culminates in the simplified equilibrium moisture ratio (X∞ /X0 ). Although the entire kinetic/equilibrium transition stage curve has
not been measured, it is estimated to be much longer in time than the kinetic stage, due to the slope found between Xdpe /X0 and
X∞ /X0 [6].
𝑋𝑡
𝑠𝑀𝑟 = (3)
𝑋0
where sMr represents the simplified moisture ratio. Xt and X0 were already previously defined, in the definition of terms of Eq. (1).
For simulation purposes, the importance of recognizing the kinetic/equilibrium transition stage, can be appreciated in Fig. 2,
where a drying curve is simulated. A retention time inside the dryer is estimated equal to tsim . The simulation would erroneously

Fig. 2. Simulation of a drying curve up to a time tsim . This graphic representation is only for explanatory purposes, since inside a real continuous
dryer, the gas temperature normally changes.

3
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

assume that the moisture at time tsim is X∞ /X0 , because the real moisture content reached in that time is Xt /X0 , since as it falls on
the kinetic/equilibrium transition stage curve. Negative values of the outlet moisture content in a simulation, using Eq. (2), would
imply that said moisture values would be located on the kinetic/equilibrium transition curve. Adjustments in the drying conditions
must be made, in order to visualize a positive value. On the other hand, if a positive value resulted, that is, above the dynamic
pseudo-equilibrium moisture content (but inside the kinetic curve), it would by no means be miscalculated. An ideal simulation for
industrial dryers would require a model taking into account the entire drying process, where the first stage should be modeled with
Eq. (2). This, however, would be impractical, and could rather be rigorously simulated, using pMr instead of Mr, together with a
model for Xdpe .
The identification of the trend of the dynamic pseudo-equilibrium moisture content, with respect to temperature, is of vital
importance in the simulation of co-current dryers, where the temperature of the exit gas decreases dramatically with respect to its
value at the entrance [9,10]. This is because the dynamic pseudo-equilibrium moisture content decreases with decreasing drying gas
[5,6], as will be seen later in the validation section.
In the pseudo-dynamic equilibrium stage, both kinetic phenomena [11] and surface thermodynamic phenomena coexist. As is well
known, in real equilibrium, the rate of condensation and evaporation of steam and liquid water, respectively, at the molecular level,
it is the same. Furthermore, this phenomenon is mainly superficial. The effects related to the kinetic phenomena decrease to the
extent that the drying process advances to the limit, where aforementioned equilibrium thermodynamic phenomena prevail. For this
reason, the dynamic pseudo-equilibrium moisture content cannot be predicted by traditional thermodynamic equations.
The numerical value of the dynamic pseudo-equilibrium moisture content (Xdpe ) decreases dramatically with increasing tempera-
ture, being predicted by exponential and potential functions [5,6]. This was to be expected since the drying curves get closer to each
other as the drying temperature increases. This separation between curves is not proportional.
Just like any other method of analysis, the conditions for determining Xdpe (a thermophysical property of drying) have been refined
in successive published experiments [5,6], and in unpublished experiments. At this point it is good to clarify that this property was
not being sought to be determined, but rather that it emerged as a scientific finding.

Goal

The objective of this method is to describe the procedure to determine the dynamic pseudo-equilibrium moisture contents in an
experimental drying kinetics study, and the calculation of the practical moisture ratio for the modeling of said experimental kinetics,
for use in rigorous design and simulation of industrial dryers.

Method

The method consists, first of all, in explaining the experimental determination of the dynamic pseudo-equilibrium moisture content
for each drying curve, determined for a material of interest. Secondly, explain how to incorporate this property in the thin layer drying
kinetics model that best fits the aforementioned curve. To convert the experimental data to data suitable for modeling, it is enough
to have a spreadsheet such as Excel®, as a computational tool. Below is the step-by-step explanation of the method.
The method is divided into four main steps. The first step is to identify that the drying process is within the dynamic pseudo-
equilibrium stage, during the measurement of an experimental drying curve. The second step is to process the experimental data in
terms of simplified moisture content vs time and identify the dynamic pseudo-equilibrium moisture content for each curve. As a third
step, the dynamic pseudo-equilibrium moisture content trend as a function of temperature must be identified and the correlation
generated. As a fourth and last step, the kinetic model is reported in terms of moisture content on a dry basis, for use in simulation.
These steps are presented schematically in Fig. 3.
In the first step, the moment in which the mass of the samples begins to vary in an order of 10 times the precision of the balance
used must be identified, during the experimental measurement of its drying curve. For example, if the precision of the balance is
0.1 mg, it should be identified when the mass of the material is changing in the order of milligrams. At this moment, the drying process
is considered to have reached the kinetic/equilibrium transition stage. Identify the mass where this kinetic/equilibrium transition
stage begins (mt ), and write down the time at which it was measured (tb ). Keep measuring the mass of the material as a function
of time, for a time between 10% and 15% additional to the time of the kinetic stage (tb x1.25), to verify the permanence in the
kinetic/equilibrium transition stage. Stop the measurements and measure the mass of dry solid (mDS ), as usual (105 °C, for 24 h, in
a laboratory oven).
In the second step, calculate the moisture content on a dry basis (Xt ) for each mass measurement or point on the kinetic curve
(mt ), as usual, using Eq. (4). Divide all moistures by X0 , to obtain sMr (Eq. (3)), and plot as a function of time (t). Using the sMr
vs t graph, identify, in each drying curve, the point where the kinetic behavior culminates, whose time (tdpe ) must be close to the
time identified in the experimental phase (tb ), and write down this simplified moisture ratio as Xdpe /X0 , one for each experimental
temperature (TG ). Validate that the equilibrium moisture contents, at the experimental temperatures, have numerical values below
the dynamic pseudo-equilibrium moisture contents.
𝑚 − 𝑚𝐷𝑆
𝑋𝑡 = 𝑡 (4)
𝑚𝐷𝑆
where Xt is the moisture content (db) of a point on the drying curve at time t (kg/kg) (water/dry matter), mt is the mass of a point
on the drying curve at time t (g), and mDS is the dry mass of the sample subjected to the measurement of its drying curve (g). The
initial masses and moistures are denoted as m0 and X0 , replacing mt and Xt in Eq. (4), respectively, when t = 0.

4
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

Fig. 3. Scheme of the dynamic pseudo-equilibrium moisture content determination method in an experimental drying kinetics study, and calculation
of the practical moisture ratio for the modeling of said experimental kinetics, for use in rigorous design and simulation of industrial dryers.

In the third step, graph Xdpe /X0 , vs TG . Model the Xdpe /X0 vs TG data, with the most efficient algebraic expression (Xdpe Function)
and measure its performance with the preferred statistics. We recommend the use of some of the statistics proposed by Faneite et al.
[5,6].
In the fourth step, calculate the drying curves in terms of pMr vs t, calculating pMr with Eq. (2). Model the pMr vs t curves, with
the most efficient thin-layer drying kinetics model, for example, as Faneite et al. [5,6]. (Kinetic model in Eq. (5)) (The modeling of
the kinetic stage of the drying curves is beyond the scope of this method).
𝑝𝑀𝑟 = 𝐾𝑖𝑛𝑒𝑡𝑖𝑐 𝑚𝑜𝑑𝑒𝑙 (5)
Equating Eqs. (2) and (5), Eq. (6) is obtained.
𝑋𝑡 − 𝑋𝑑𝑝𝑒
= 𝐾𝑖𝑛𝑒𝑡𝑖𝑐 𝑚𝑜𝑑𝑒𝑙 (6)
𝑋0 − 𝑋𝑑𝑝𝑒
Dividing the left side of Eq. (6) by X0 , Eq. (7) is obtained.
𝑋𝑡 ∕𝑋0 − 𝑋𝑑𝑝𝑒 ∕𝑋0
= 𝐾𝑖𝑛𝑒𝑡𝑖𝑐 𝑚𝑜𝑑𝑒𝑙 (7)
𝑋0 ∕𝑋0 − 𝑋𝑑𝑝𝑒 ∕𝑋0
Simplifying X0 /X0 = 1 and substituting Xdpe /X0 for Xdpe function in Eq. (7), Eq. (8) is obtained.
𝑋𝑡 ∕𝑋0 − 𝑋𝑑𝑝𝑒 𝑓 𝑢𝑛𝑐𝑡𝑖𝑜𝑛
= 𝐾𝑖𝑛𝑒𝑡𝑖𝑐 𝑚𝑜𝑑𝑒𝑙 (8)
1 − 𝑋𝑑𝑝𝑒 𝑓 𝑢𝑛𝑐𝑡𝑖𝑜𝑛
Rearranging Eq. (8) step by step, Eqs. (9), (10), and (11) are obtained successively.
( )
𝑋𝑡 ∕𝑋0 − 𝑋𝑑𝑝𝑒 𝑓 𝑢𝑛𝑐𝑡𝑖𝑜𝑛 = 1 − 𝑋𝑑𝑝𝑒 𝑓 𝑢𝑛𝑐𝑡𝑖𝑜𝑛 ⋅ (𝐾𝑖𝑛𝑒𝑡𝑖𝑐 𝑚𝑜𝑑𝑒𝑙) (9)

( )
𝑋𝑡 ∕𝑋0 = 𝑋𝑑𝑝𝑒 𝑓 𝑢𝑛𝑐𝑡𝑖𝑜𝑛 + 1 − 𝑋𝑑𝑝𝑒 𝑓 𝑢𝑛𝑐𝑡𝑖𝑜𝑛 ⋅ (𝐾𝑖𝑛𝑒𝑡𝑖𝑐 𝑚𝑜𝑑𝑒𝑙) (10)

[ ( ) ]
𝑋𝑡 = 𝑋0 ⋅ 𝑋𝑑𝑝𝑒 𝑓 𝑢𝑛𝑐𝑡𝑖𝑜𝑛 + 1 − 𝑋𝑑𝑝𝑒 𝑓 𝑢𝑛𝑐𝑡𝑖𝑜𝑛 ⋅ (𝐾𝑖𝑛𝑒𝑡𝑖𝑐 𝑚𝑜𝑑𝑒𝑙) (11)
where Xdpe function is the correlation of the dynamic pseudo-equilibrium moisture content as a function of the temperature of the
drying gas in terms of the simplified moisture ratio, Kinetic model is the thin layer drying model that best fits the experimental drying

5
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

data in terms of the practical moisture ratio. Xt , X0 , Xdpe , and pMr were already previously defined, in the definition of terms of
Eqs. (1) and (2).
Eq. (11) is the Kinetic model in terms of moisture content on a dry basis, based on dynamic pseudo-equilibrium moisture content.

Applications

Eq. (11) applies to the first finite volume, where the input moisture content corresponds to the initial moisture of the material
to be dried. In the simulation of industrial dryers, the equipment is divided, in a virtual way, into volumes of finite size (hence the
name of the technique), for its rigorous design. In each of these volumes a balance of mass, energy and movement of matter through
the dryer is established [9]. The output of matter and energy from a previous finite volume is the input of matter and energy from
the next finite volume (this does not apply, of course, to lost heat) [9]. The material and energy output of the last finite volume is
the output of the dryer [9]. For the subsequent finite volumes, Eq. (11) must be adjusted. The adjustment of Eq. (11), for the rest of
the finite volumes, is beyond the scope of this method. Eq. (11) also applies to batch dryers that work with a fixed gas temperature.

Validation of the method

Two cases of experimental thin-layer drying kinetics are presented below, where the method described in the previous section was
applied, and where its validity is demonstrated. These two cases correspond to the first two published studies, of a group of more than
ten studies, each related to a material of biological origin of interest, for the areas of biomass, biorefining, postharvest processing, and
animal feed. As previously mentioned, the methodology established from the first of these studies, whose objective was to generate
inputs for the simulation and rigorous design of industrial dryers, allowed us to visualize this finding that was not being sought. That
same stated objective allowed us to discover the importance of dynamic pseudo-equilibrium moisture for the ultimate use of drying
kinetics experimental data. Finally, the strength of the research group in the areas of phase equilibrium thermodynamics, transport
phenomena and surface phenomena, typical areas of Chemical Engineering, finished giving theoretical foundation to the finding, and
differentiating real equilibrium from pseudo-equilibrium. The drying kinetics data, presented in the following validation cases, was
processed in Excel○R .

Case 1. Green plantain (Musa paradisiaca L.) peel

The thin-layer drying kinetics of green plantain (Musa paradisiaca L.) peel was published by Faneite et al. [5], in the temperature
range from 60 °C to 110 °C. This was the first work published, where the method was applied. Table 1 shows the final experimental
points of the drying curves, where the difference in masses is included, point by point, with respect to the previous point. Mass
differences in the order of milligrams, 10 times the precision of the balance (of 0.1 mg), indicate the end of the experiment, as
established by this method.
As a next step, the complete experimental data is processed in Excel® to obtain the drying curves and identify the dynamic
pseudo-equilibrium moisture content. The drying curves for the green plantain peel, from 60 °C to 110 °C, in terms of the simplified
moisture ratio (Xt /X0 ), within the time range of interest, can be seen in Fig. 4. Kinetic and pseudo-dynamic equilibrium stages, as well
as pseudo-dynamic equilibrium moisture content were indicated with markers of different colors. It is possible to clearly appreciate
the change in the behavior of the curve, towards a more horizontal one.
As a last step, the trend of the experimental points of the dynamic pseudo-equilibrium moisture content is identified and the
corresponding mathematical model is established. This model will be included in the Kinetic model in terms of moisture content
on a dry basis. The Xdpe /X0 trend is potential, both in terms of time and as a function of temperature (Figs. 4 and 5, respectively),
coinciding with the hypothesis expressed in the Background. Eq. (12) was the model that best fits the behavior of Xdpe /X0 as a function

Table 1
Period of the drying curve, expressed in average mass of the replicates, where the change from kinetic behavior to the kinetic/equilibrium transition
stage occurs for the green plantain (Musa paradisiaca L.) peel at different temperatures.

60 °C 70 °C 80 °C 90 °C 100 °C 110 °C

t m DIFF t m DIFF t m DIFF t m DIFF t m DIFF t m DIFF

(min) (g) (mg) (min) (g) (mg) (min) (g) (mg) (min) (g) (mg) (min) (g) (mg) (min) (g) (mg)

320 2.7589 −15.07 265 2.7157 −17.87 185 2.4417 −14.80 150 2.6548 −18.73 138 2.7246 −21.23 117 2.8026 −14.73
325 2.7443 −14.60 270 2.7074 −8.30 190 2.4294 −12.37 153 2.6368 −17.93 141 2.7022 −22.40 120 2.7856 −17.00
330 2.7338 −10.47 275 2.6944 −13.07 195 2.4192 −10.13 156 2.6248 −12.07 144 2.6879 −14.30 123 2.7739 −11.73
335 2.7323 −1.50 280 2.6870 −7.33 200 2.4111 −8.17 159 2.6175 −7.30 147 2.6821 −5.80 126 2.7654 −8.47
340 2.7288 −3.57 285 2.6776 −9.47 205 2.4046 −6.47 162 2.6137 −3.80 150 2.6748 −7.33 129 2.7640 −1.40
345 2.7225 −6.30 290 2.6715 −6.07 210 2.3997 −4.87 165 2.6125 −1.20 153 2.6765 1.77 132 2.7613 −2.67
350 2.7187 −3.80 295 2.6688 −2.67 215 2.3940 −5.73 168 2.6068 −5.63 156 2.6706 −5.97 135 2.7572 −4.17

t: time; m: mass (measured with analytical balance with a precision of ± 0.1 mg); DIFF: difference in masses, previous point minus the current one.
Bold: mass where the change in mass began to have an order of 10 times the precision of the balance. Source: Faneite et al. [5].

6
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

Fig. 4. End of the experimental drying curves of green plantain (Musa paradisiaca L.) peel at temperatures from 60 °C to 110 °C [5], from 10 to 10
°Celsius, from right to left. Experimental points: kinetic stage ( ), dynamic pseudo-equilibrium moisture content ( ), dynamic pseudo-equilibrium
stage ( ). Xdpe /X0 tendency ( ).

Fig. 5. Simplified moisture ratio of dpe moisture content (Xdpe /X0 ) as a function of absolute temperature (K) for green plantain (Musa paradisiaca
L.) peel. The markers represent the experimental points. The continuous line represents Eq. (12).

of temperature, for this material.

𝑋𝑑𝑝𝑒 ∕𝑋0 = 5.12656418274477E21 ⋅ (𝑇 + 273.15)−9.16792923 (12)

where Xdpe /X0 is the simplified moisture ratio at the beginning of the dynamic pseudo-equilibrium stage at the drying gas temperature
T, and T is the temperature of the drying gas ( °C).
The performance of Eq. (12) was excellent. A very high Coefficient of Determination (R2 ), of 99.48%, indicates how closely the
model fits experimental behavior, where 100 % is a perfect fit. A very low Mean Bias Error (MBE·10,000), of 0.14, indicates how
evenly the model errors are distributed around 0, where 0 is a perfect distribution. A very low Root Mean Square Error (RMSE·10,000),
of 6.8, indicates the bias with which the model makes a prediction, where 0 indicates no bias at all. A Standard Error of Estimation
(SEE·10,000) of 8.4, of the same order as RMSE, indicates that the number of parameters of the model is not exaggerated.
Dynamic pseudo-equilibrium moisture contents could not be compared with equilibrium moisture contents for green plantain
peels. It only exists in the literature, a data reported by Kappaun et al. [7], as equilibrium moisture content. They erroneously
established a value of 6.0 kg/kg ± 0.3 kg/kg (water/dry matter) for 40 °C, 60 °C, and 80 °C, but what they really got were average
values for a portion of the kinetic/equilibrium transition stage at the temperatures mentioned. Finally, the modeling of the thin-layer
drying kinetics of the green plantain peels can be reviewed in the work of Faneite et al. [5].

Case 2. Peeled cassava root slices

The kinetics of thin-layer drying of cassava root, peeled and cut into 5 mm and 6 mm slices, was published by Faneite et al. [6],
in the temperature range from 55 °C to 105 °C. This was the second work published, where the method was applied. They found the
times in which the kinetic behavior culminated for each drying curve, as was done in the previous case, presenting the results in
Tables 2 and 3, respectively (bold numbers). Unlike the case of the drying data of green plantain (Musa paradisiaca L.) peel, reported
in Table 1, the average of the masses of Tables 2 and 3, does not generate four endpoints with a difference in milligrams in all cases.
However, in the replicas, this requirement was visualized in full compliance. This is due to slight differences in the initial masses of
the replicates. Likewise, the identification of these points in tables 2 and 3 is purely referential since the final decision is made by
looking at the drying curves on a graph.
Drying curves for cassava root, peeled and cut into 5 mm and 6 mm slices, at the temperatures studied and within the time range
of interest (in terms of the simplified moisture ratio) are shown in Figs. 6(A) and (B), respectively. The behavior or shape of the drying
curve at the end of it coincides with that of the drying of green banana peels. As in Fig. 3, kinetic and pseudo-dynamic equilibrium

7
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

Table 2
Period of the drying curve, expressed in average mass of the replicates, where the change from kinetic behavior to the kinetic/equilibrium transition
stage occurs for peeled cassava root, sliced 5 mm thick at different temperatures.

55 °C 65 °C 75 °C 85 °C 95 °C 105 °C

t m DIFF t m DIFF t m DIFF t m DIFF t m DIFF t m DIFF

(min) (g) (mg) (min) (g) (mg) (min) (g) (mg) (min) (g) (mg) (min) (g) (mg) (min) (g) (mg)

340 14.7395 72.03 290 14.2269 57.20 285 11.2697 15.77 235 12.0896 23.30 177 10.5685 18.90 132 12.0013 29.53
345 14.6467 92.83 295 14.1833 43.57 290 11.2555 14.17 240 12.0700 19.60 180 10.5531 15.43 135 11.9822 19.17
350 14.6223 24.37 300 14.1425 40.80 295 11.2350 20.47 245 12.0549 15.10 183 10.5399 13.17 138 11.9626 19.57
355 14.6084 13.90 305 14.1037 38.80 300 11.2211 13.93 250 12.0472 7.63 186 10.5302 9.77 141 11.9561 6.47
360 14.5972 11.23 310 14.1035 0.27 305 11.2083 12.83 255 12.0404 6.83 189 10.5188 11.33 144 11.9390 17.10
365 14.5937 3.50 315 14.1032 0.23 310 11.1976 10.63 260 12.0347 5.67 192 10.5108 8.03 147 11.9124 26.67
370 14.5904 3.33 320 14.1030 0.27 315 11.1953 2.30 265 12.0293 5.40 195 10.5067 4.13 150 11.8904 22.00

t: time; m: mass (measures with analytical balance with a precision of 0.1 mg); DIFF: difference in masses, previous point minus the current one.
Bold: mass where the change in mass began to have an order of 10 times the precision of the balance. Source: Faneite et al. [6].

Table 3
Period of the drying curve, expressed in average mass of the replicates, where the change from kinetic behavior to the kinetic/equilibrium transition
stage occurs for peeled cassava root, sliced 6 mm thick at different temperatures.

55 °C 65 °C 75 °C 85 °C 95 °C 105 °C

t m DIFF t m DIFF t m DIFF t m DIFF t m DIFF t m DIFF

(min) (g) (mg) (min) (g) (mg) (min) (g) (mg) (min) (g) (mg) (min) (g) (mg) (min) (g) (mg)

370 14.3818 103.63 330 14.5985 19.13 295 11.8951 43.57 260 11.3468 44.03 219 12.4343 16.66 189 12.1828 25.33
375 14.3062 75.57 335 14.5892 9.27 300 11.8543 40.80 265 11.3023 44.43 222 12.4004 33.86 192 12.1491 33.77
380 14.2953 10.97 340 14.5172 72.03 305 11.8155 38.80 270 11.2758 26.57 225 12.3667 33.69 195 12.1204 28.71
385 14.2713 23.97 345 14.4243 92.83 310 11.8153 0.27 275 11.2539 21.90 228 12.3577 9.05 198 12.1153 5.07
390 14.2643 7.00 350 14.4000 24.37 315 11.8150 0.23 280 11.2434 10.43 231 12.3486 9.05 201 12.1119 3.38
395 14.2624 1.87 355 14.3861 13.90 320 11.8148 0.27 285 11.2360 7.40 234 12.3396 9.05 204 12.1102 1.69
400 14.2612 1.27 360 14.3748 11.23 325 11.8146 0.20 290 11.2306 5.40 237 12.3305 9.05 207 12.1085 1.69

t: time; m: mass (measures with analytical balance with a precision of 0.1 mg); DIFF: difference in masses, previous point minus the current one.
Bold: mass where the change in mass began to have an order of 10 times the precision of the balance. Source: Faneite et al. [6].

Fig. 6. End of the experimental drying curves of peeled cassava roots ( ), (A) sliced 5 mm thick, and (B) sliced 6 mm thick, at temperatures from
55 °C to 105 °C [6], from 10 to 10 °Celsius, from right to left. Experimental points: kinetic stage ( ), dynamic pseudo-equilibrium moisture content
( ), dynamic pseudo-equilibrium stage ( ). Xdpe /X0 tendency ( ).

8
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

Fig. 7. Simplified moisture ratio of dpe moisture content as a function of absolute temperature for peeled cassava root, ( ) sliced 5 mm thick, and
( ) sliced 6 mm thick. The solid line represents Eq. (13).

stages, as well as pseudo-dynamic equilibrium moisture content were indicated with markers of different colors. The trend of Xdpe /X0
is exponential in both cases.
The behavior of the Xdpe /X0 as a function of the absolute temperature (Fig. 7) responds to an exponential behavior, and the
prediction model is represented by Eq. (13).

𝑋𝑑𝑝𝑒 ∕𝑋0 = (−8.356210 ⋅ 𝐿 + 58.268958) ⋅ 𝑒(0.002498⋅𝐿 − 0.029209)⋅(𝑇 +273.15) (13)

where L is the layer height (mm). Xdpe /X0 and T were already previously defined, in the definition of terms of Eq. (12).
The capacity of the Eq. (13) to reproduce the trend of the experimental data was excellent, being the results R2 for 5 mm and
6 mm, of 97.98% and 96.76%, respectively. The capacity of the model to distribute the errors around 0 was excellent too, obtaining
0.658 and 0.813, respectively, in terms of MBE·10,000. The values of RMSE·10,000 indicated a low bias between the model and the
experimental data of 19.0 and 23.8, respectively. The values of SEE·10,000 were in the same order and with numerical values quite
close to the RMSE·10,000, obtaining 21.9 and 27.5, respectively. This last result indicate that the number of parameters of this model
is not exaggerated, since it does not significantly affect the RMSE bias.
The fact that the layer height influences the numerical value of Xdpe implies that transport phenomena prevail over surface
phenomena. Likewise, despite the fact that in this transition stage, both transport phenomena (typical of kinetics) and those typical
of thermodynamic equilibrium coexist, in the Xdpe limit, the first prevail, due to its proximity to this first stage.
Once the Xdpe /X0 model has been obtained, as a function of temperature, and its performance has been measured, it is verified
whether, indeed, the equilibrium moisture content is below the dynamic pseudo-equilibrium moisture content. For this, the Xdpe /X0
values are multiplied by the initial moisture content on an average dry basis, of each curve, and the Xdpe values are obtained, which
can be seen in Table 4. Likewise, using the proposed model by Moreno et al. [12], with parameters calculated from the data of Best
[13], and using the environmental and drying chamber conditions from the experiments of Faneite et al. [6], the equilibrium moisture
contents were obtained, which are also shown in Table 4. It can be seen that the dynamic pseudo-equilibrium moisture content is
higher in all temperature and layer height conditions than the true equilibrium moisture content.
The relative difference (RD) of the equilibrium moisture content with respect to the dynamic pseudo-equilibrium moisture content
can be calculated with Eq. (14).
( )
𝑅𝐷 = 𝑋𝑑𝑝𝑒 − 𝑋∞ ⋅ 100%∕𝑋∞ (14)

Table 4
Comparison of equilibrium and dynamic pseudo-equilibrium moisture contents of peeled cassava root in 5 mm and 6 mm slices
at conditions reported by Faneite et al. [6].

Drying kinetic experiments at 5 mm layer height Drying kinetic experiments at 6 mm layer height

TDC ( °C) RHA (%) RHDC (%) X∞ (kg/kg) Xdpe (kg/kg) TDC ( °C) RHA (%) RHDC (%) X∞ (kg/kg) Xdpe (kg/kg)

55 71.61 10.65 0.0620 0.0918 55 62.28 15.25 0.0654 0.1179

65 68.50 9.89 0.0575 0.0870 65 68.50 9.89 0.0575 0.1047
75 69.33 4.96 0.0505 0.0810 75 68.50 6.43 0.0515 0.0985
85 73.14 3.38 0.0463 0.0667 85 64.89 4.04 0.0468 0.0783
95 72.38 2.13 0.0427 0.0563 95 65.30 2.88 0.0432 0.0688
105 72.38 0.00 0.0388 0.0522 105 68.80 0.00 0.0388 0.0564

TDC : Drying chamber temperature. RHA : Ambient relative humidity. RHDC : Relative humidity inside the drying chamber. X∞ :
Equilibrium moisture content (kg/kg) (water/dry solid) calculated with the procedure suggested by Moreno et al. [12], using the
parameters obtained by correlating the data from Best [13], and the conditions of the drying chamber reported by Faneite et al.
[6]. Xdpe : Dynamic pseudo-equilibrium moisture content (kg/kg) (water/dry solid) obtained from Faneite et al. [6].

9
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

Fig. 8. The relative difference of the equilibrium moisture (RD) with respect to the dynamic pseudo-equilibrium moisture as a function of temper-
ature at 5 mm and 6 mm layer height.

where RD is the relative difference (%) between the equilibrium moisture content (X∞ in kg/kg (water/dry matter)), and the dy-
namic pseudo-equilibrium moisture content Xdpe in kg/kg (water/dry matter)), measured or estimated at the same drying conditions
(temperature and relative humidity of the drying gas).
Fig. 8 is obtained by plotting DR. as a function of temperature for each layer height. It can be seen in both cases, and with an
almost identical behavior, that when the drying temperature rises from 55 °C to 105 °C, RD increases up to 75 °C, and then decrease.
An in-depth analysis on RD behavior is beyond the scope of this validation. Finally, the modeling of the drying kinetics in a thin-layer
of the cassava root, peeled and cut into 5 mm and 6 mm slices, at a temperature range between 55 °C and 105 °C, can be reviewed
in the work of Faneite et al. [6].

Conclusion

Both for the drying of the green banana peel and for the cassava root, peeled and cut into 5 mm and 6 mm slices, a limit was found
in the moisture content, where the initial kinetic behavior of monotonic decrease in moisture content, changes to a more horizontal
trend. Said limit, baptized as the dynamic pseudo-equilibrium moisture content, is below the typical storage moisture content (around
10%, on a wet basis), but above the real equilibrium moisture content, so this latter moisture should not be used to calculate the
practical moisture ratio. The practical moisture ratio (calculated with the dynamic pseudo-equilibrium moisture content instead of
the real equilibrium moisture content) is the correct expression for the moisture content used as a dependent variable in the kinetic
models for simulation and rigorous design of industrial dryers.
There are no data published in the literature about the equilibrium moisture content of the green plantain peels, for comparative
effects with the dynamic pseudo-equilibrium moisture content. However, dynamic pseudo-equilibrium moisture contents are higher
than the equilibrium moisture contents, for all temperatures, of the drying of the cassava root, peeled and cut into 5 mm and 6 mm
slices. The dynamic pseudo-equilibrium moisture content of cassava root drying at 6 mm layer height was above that at 5 mm layer
height. These results confirm the existence of a transition stage between the initial kinetic stage (governed by transport phenomena)
and the final state of thermodynamic equilibrium (governed by surface phenomena), where both phenomena are conjugated.

CRediT author statement

Alexis Manuel Faneite Noguera: Finding, Conceptualization, Methodology, and Validation. Ignacio Angós Iturgaiz: Concep-
tualization.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Data availability

Data will be made available on request.

Acknowledgments

The author thanks Professor Alexis Ferrer Ocando (University of Zulia, Venezuela), and Professor Paloma Vírseda (Public University
of Navarra, Spain), for facilitating all the experiments that led to the discovery of the transition period between the kinetic drying
process and the thermodynamic equilibrium of drying.
This research received partial funding of Industrial Dryers of Venezuela Corporation (SECAVENCA) and Open Access funding
provided by Universidad Pública de Navarra (UPNA)

10
A.M. Faneite Noguera and I. Angós Iturgaiz MethodsX 11 (2023) 102410

References

[1] A. Smith, Moisture ratio, Science 48 (1228) (1918) 48–49, doi:10.1126/science.48.1228.48.

[2] W.K. Lewis, The rate of drying of solid materials, Ind. Eng. Chem. 13 (5) (1921) 427–432 https://pubs.acs.org/doi/pdf/10.1021/ie50137a021.
[3] F. Freire, A. Figueiredo, P. Ferrao, Modelling High Temperature, Thin-layer, Drying Kinetics of Olive Bagasse, PH-Postharvest Technology, J. Agric. Eng. Res. 78
(4) (2001) 397–406, doi:10.1006/jaer.2000.0657.
[4] H. Li, H.S. Ramaswamy, Osmotic dehydration: dynamics of equilibrium and pseudo-equilibrium kinetics, Int. J. Food Prop. 13 (2010) 234–250,
doi:10.1080/10942910802331272.
[5] A. Faneite, A. Rincón, A. Ferrer, I. Angós, G. Arguello, Mathematical modeling of thin-layer drying of green plantain (Musa paradisiaca L.) peel, Int, Food Res. J.
23 (5) (2016) 2088–2095 URL: http://www.ifrj.upm.edu.my/23/20(05)/202016/(33).pdf .
[6] A. Faneite, J.L. Parra, W. Colón, A. Ferrer, I. Angós, G. Argüello, New thin-layer drying models for the design and simulation of cas-
sava root dryers and phenomenological study of interaction water-starch during diffusion, Int. Food Res. J. 27 (1) (2020) 182–196 URL:
http://www.ifrj.upm.edu.my/27/20(01)/202020/19/20-/20IFRJ161084.R1.pdf .
[7] M.C. Kappaun R, C.E. Demaman O, B.M. Saorin P, M.S. Wilk W, I.A. Fernandes, R.L. Cansian, G.T. Backes, A. Junges, Potential use of green
banana peel waste: modeling of drying and determination of physicochemical and antioxidant properties, Biomass Conv. Bioref. (2022) 1–12
https://link.springer.com/article/10.1007/s13399-022-03511-z.
[8] K. Molnár, Experimental Techniques in Drying, in: A. S. Mujumdar (Ed.), Handbook of industrial drying. Third edition, Boca Ratón: CRC Press, 2006, pp. 39–40.
https://doi.org/10.1201/9781420017618.
[9] A. Iguaz, A. Esnoz, G. Martı́nez, A. López, P. Vírseda, Mathematical modelling and simulation for the drying process of vegetable wholesale by-products in a
rotary dryer, J. Food Eng. 59 (2-3) (2003) 151–160, doi:10.1016/S0260-8774(02)00451-X.
[10] A. Faneite, A. Ferrer, J. Tudare, R. Cardozo, Biorefinación de la Lemna obscura del Lago de Maracaibo. Parte I. Acondicionamiento de Materia Prima, in:
M. Boves, J. Rincón (Eds.), Eutrofización Del Lago de Maracaibo: Pasado, Presente y Perspectivas. First edition, Universidad del Zulia, 2013, pp. 225–255.
doi:10.13140/2.1.4474.1449
[11] R.E. Treybal, Mass-Transfer Operations. Second Edition, McGraw-Hill, 1968, pp. 747–753.
[12] F.L. Moreno, A. Parra-Coronado, J.H. Camacho-Tamayo, Mathematical simulation parameters for drying of cassava starch pellets, Eng. Agríc. 34 (6) (2014)
1234–1244 URL: https://www.scielo.br/j/eagri/a/Z6CGGvLzgZ5zpSnRb4JynNc/?format=pdf&lang=en .
[13] R. Best, Cassava processing for animal feed, in E. J. Weber, J. H. Cock, A. Chouinard (Eds), Cassava Harvesting and Processing, Proceedings of a Workshop Held
At CIAT, Cali, Colombia, from April 24 to 28, 1978. URL: https://idl-bnc-idrc.dspacedirect.org/bitstream/handle/10625/20447/IDL-20447.pdf?sequence=1

11