Вы находитесь на странице: 1из 7

Lecture-16: The ideal plug flow reactor

The next idealized reactor configuration we will consider is the ideal plug flow tubular reactor.
In this model, we assume that the reactor is a tube in which the velocity is in the axial direction
only. We further assume that the velocity and all other quantities are constant across the tube
diameter, and that diffusion and conduction in the axial direction are negligible. With these
assumptions, the species mole balance equations are

If the velocity is independent of axial position (which will be the case if density is constant), then
it can be taken out of the derivative to get

Defining a residence time τ (the time a fluid element has spent in the reactor when it reaches
position x) by

this equation becomes

which is identical to the constant volume batch reactor equation with the clock time t replaced by
the residence time τ.

If the density in the reactor cannot be assumed to be constant (because of changes in the number
of moles or changes in the temperature of a gas, for example) then the plug flow reactor balances
are not identical to those for the batch reactor. However, we can still define the residence time in
a similar way as

where vx,o is the axial velocity at the inlet. Then the equation becomes

The velocity at a given point in the reactor can be related to the density at that point using the
total mass balance. For the steady-state plug flow reactor, the mass balance reduces to
which can directly be integrated between the inlet (where the velocity is vx,o and the density is ρo)
and some other point in the reactor to give
ρvx = ρovx,o

The local velocity can often be computed from the ideal gas law using the local composition and
temperature. For isothermal reaction among gases, it can be written in terms of the extent of
reaction (for as many linearly independent reactions as we have) and the overall change in mole
number for those reactions. For a mixture of ideal gases, the velocity is

where T is the temperature, n/no is the ratio of the number of moles at a particular position to the
number of moles in the feed, and p is the pressure.

The most direct measure of the reactor’s capability to carry out the reaction is given by the total
residence time based on the inlet flow rate and total reactor volume. Froment and Bischoff
designate this as θ, but most other authors use the symbol τ.

Where L is the reactor length, V is the reactor volume, and Qo is the volumetric flow rate of
reactant to the reactor. In practice, reactors are often described in terms of the reciprocal of this,
which is usually called ‘space velocity’. For catalytic reactions, it is not the volume of the reactor
that matters, but the amount of catalyst, so in cases like that, the 'space velocity' might be
replaced by a term like 'weight-hourly-space-velocity' (WHSV), which is usually defined as the
mass flow rate of reactant divided by the mass of catalyst in the reactor, so that it has units of
inverse time (hr -1 for an hourly space velocity).

The corresponding energy balance equation for the ideal PFTR is

where Rt is the tube radius, U is an overall heat transfer coefficient for heat transfer through the
tube wall to a heat exchange fluid at temperature Tc. Of course this term for heat flux into the
reactor through the walls could take many other forms.

Introducing the residence time as before and dividing through by the density and specific heat
gives
This is the basic energy balance for an ideal PFTR. It is almost identical to the balance for the
batch reactor – the only difference is in the term describing heat transfer into or out of the
reactor. Note that the inlet density is used in this equation, and that it is valid for non-constant
density (as well as constant density) reactors.

Because of the similarity between the equations describing the plug flow reactor and the
equations that describe the batch reactor, much of what we said about the batch reactor also
applies here. It is straightforward to integrate the material and energy balance equations given
above for a particular reactor configuration. Although we won’t discuss them here, you should
be able to use these equations to solve simple design problems like determining the reactor
volume required to process a specified amount of reactant to a specified conversion, etc. For
cases where the equations cannot be integrated analytically, you can use the same methods for
initial value problems that we have previously used for the batch reactor (since they are the same
equations).

For an adiabatic plug flow reactor with a single overall reaction, there is a simple relationship
between the reactor temperature and the reactant concentration, just as there was for the batch
reactor. The equations for a single reaction in an adiabatic, constant density PFTR (in terms of
the concentration and production rate of the reactant A) are

From these, we see that

And this may be integrated to give the same relationship that we had for the adiabatic batch
reactor with a single reaction

Thus, the adiabatic reaction temperature that we defined in the batch reactor has the same
definition and value in the plug flow reactor.

Optimal Operation of a PFTR:


Obtaining the optimal single temperature or temperature profile in the PFTR is also very similar
to what we found for the batch reactor. If we plan to operate the reactor isothermally (at a single
temperature) then there may be an optimum temperature to select. For a single, irreversible
reaction, the optimal temperature is simply the highest possible temperature, since that will
provide the highest possible rate. This temperature will be determined by other factors, like
boiling of the reactants or operational limits of the reactor. For an endothermic, reversible
reaction, the best temperature is again the highest possible temperature, since both the
equilibrium conversion and the reaction rate will increase with temperature. The only case where
there may be an optimum temperature for a single reaction is when we have an exothermic,
reversible reaction. For such a reaction, the rate will increase with temperature, but the
equilibrium conversion will decrease with increasing temperature. Just as in the batch reactor,
the optimal temperature profile for a single reaction is the one that optimizes the reaction rate at
every point in the reactor. The discussion for the batch reactor has been cut, pasted, and very
slightly modified here.

For the single reversible reaction, the rate is

Taking the partial derivative of this with respect to T at fixed concentrations gives

Setting this equal to zero and multiplying through by RT2/A gives

This can be rearranged to give

or

This gives the optimal temperature profile in terms of the concentrations (or the extent of
reaction or whatever reaction progress variable we choose). Initially, for an exothermic reaction,
this will give an optimal temperature that is negative (not a physically significant result). The
optimal temperature becomes infinite when the quantity inside the logarithm in the denominator
is equal to one. That is

For conversions smaller than this critical value, the optimal temperature is ‘as high as possible’.
As reaction proceeds past this point, the optimal temperature decreases. So, the optimal
temperature profile for a PFTR with a single, reversible reaction is to run the reactor isothermally
at the highest temperature possible (Tmax) until the conversion reaches a point where the optimal
temperature obtained from the equation above is equal to Tmax. After that point, the temperature
should decrease along the length of the reactor, with the optimum value determined from the
equation above. We could illustrate this using the same contour plot of reaction rate vs.
temperature and extent of reaction that we used for the batch reactor in the previous set of notes.
We will come back to this and consider it in more detail for plug flow reactors when we discuss
fixed bed catalytic reactors in a few days.

As was the case for the batch reactor, in general we will need to write an objective function that
describes what we want to optimize. This may include conversion, yield, selectivity, and other
variables. Given that objective function, we can minimize (or maximize) it by varying whatever
design or operating parameters we have available. An example would be an objective function
that approximates the net profit in terms of reactor outlet composition, energy input to the
reactor, etc. We might be able to vary the reactor size and heat exchange fluid temperature and
flow rate. We would use the same optimization methods that we have used in the past to
maximize the profit (analogous to minimizing the sum of the squares of the errors) by varying
the fluid temperature, fluid flowrate, and reactor volume (analogous to varying the rate
constants).

Time-Dependent Operation of the PFTR:


Before leaving the ideal plug flow reactor, we will briefly consider its time-dependent operation.
For the non-steady-state PFTR, the governing equations are

These are 1st order differential equations. Mathematically, they are considered hyperbolic partial
differential equations. They can be solved using the method of characteristics. Any significant
discussion of that would go beyond the scope of this course. This type of equation leads to
solutions that change via the propagation of waves that move with a characteristic speed. In this
case, that speed is simply the fluid velocity. The same type of equations arises in gas dynamics
(high speed, inviscid flows), where the characteristic velocity is the speed of sound. The partial
differential equations above can be transformed using a new independent coordinate defined by

Then

and

Substituting these into the governing equations leaves us with


So, along the characteristics defined by ξ = ½(x - vxt), the net rate of all of the reactions is zero.
Likewise, along the characteristics defined by ζ = ½(x + vxt), we would find that

Along these characteristics, we have the same equations that we had for the steady state PFTR.

Physically, this behavior arises from the fact that fluid elements do not interact as they move
down the reactor. We can imagine each individual fluid element as a batch reactor with a batch
time equal to the reactor residence time. Any changes in the reactor inlet conditions have no
effect on fluid that is already in the reactor, and instantaneously affect the fluid that enters the
reactor after the change is made. The change then propagates down the reactor with the same
velocity that the fluid moves down the reactor. These equations indicate that the concentration
and temperature profiles for the adiabatic reactor are uniquely determined by the inlet conditions.
If there is heat addition or removal at the walls it is possible to have multiple steady-state
solutions to the equations.

Even for the adiabatic reactor, where the solution is determined uniquely by the inlet conditions,
we could consider the parametric sensitivity of the solution to the initial conditions. There may
be conditions where a very slight change in the inlet conditions leads to a large change in the
reactor conditions. This is often true for an exothermic reaction. For sufficiently low inlet
temperature, very little reaction will occur. Near some critical temperature, sufficient reaction
will occur at the inlet temperature to raise the temperature enough to cause a substantial increase
in the reaction rate. The process is autocatalytic in nature. The more reaction that occurs, the
faster the reaction rate becomes. This can lead to a sharp change from an inlet temperature where
almost no reaction occurs to one where there is almost complete conversion of reactant.

We will consider these questions of stability, parametric sensitivity, and time-dependent


behavior in much greater detail for the continuous stirred tank reactor (CSTR) that we will study
next. In that case, the mathematics are simpler and the likelihood of physically realizable
multiple steady state conditions is greater. We will consider other issues related to tubular
reactors such as axial dispersion, pressure drop, radial non-uniformity, etc. later when we look at
the design of fixed bed catalytic reactors. This widely used reactor type is often treated as an
ideal PFTR.
After completing your study of these lecture notes you should be able to:

(1) Understand, derive, and use mass, energy, and species balance equations for the ideal PFTR
with constant or non-constant fluid density
(2) Solve (numerically) the balance equations for a non-isothermal PFTR to obtain
concentrations and temperature in the reactor as functions of position or residence time
(3) Find the reactor length, reactor volume, or residence time required to achieve a particular
concentration or extent of reaction in a PFTR
(4) Find the optimal temperature or temperature trajectory for a PFTR with a single reaction
(5) Derive and use the equations describing unsteady-state operation of a PFTR.

Вам также может понравиться