COMMISSIONER'S DECISION
Section 2: OPTIMIZING PERFORMANCE of a Multi-Unit Power Plant
Optimizing the performance of a multi unit power plant which produces steam
energy from a plurality of different fuels where the input perturbations to
the system affecting its performance cannot be directly and accurately meas-
ured is statutory subject matter.
Final Action : Reversed.
***********
Patent application 273,956 (Class 341-110), was filed on March 15, 1977
for an invention entitled "METHOD OF OPTIMIZING THE PERFORMANCE OF A
MULTI-UNIT POWER PLANT". The inventors are Louis S. Adler et al,
(assignors to Measurex Corporation). The Examiner in charge of the appli-
cation took a Final Action refusing to allow it to proceed to patent.
The subject matter of this application relates to optimizing the perform-
ance of a multi-unit power plant by determining the incremented efficiency
of the boilers and thereafter determining the index of performance and
reallocating the sequence of their use in the most optimum manner. It is
concerned with power plants in the wood pulp industry where the boilers
are fired with a base fuel such as coal or wood chips and a swing fuel
such as oil. The steam produced is used for several purposes such as
generating electricity, heating reactors and drying pulp. In this applica-
tion the applicant determines changes in efficiency resulting from small
incremental changes in fuel used, and from that, coupled with the cost
of the fuels then calculates the most effective allocation of fuels to be
used for a particular combination of desired uses.
In the Final Action the examiner rejected the disclosure and claims as
being essentially "directed to an algorithm for controlled power plant
parameters and therefore unpatentable subject matter under Section 2."
That action stated (in part):
...
It is well known in the systems art to employ computers for
feedback control ie., sensing flow rates etc. and the applic-
ant states the same on page 6 lines 16-17. The equations as
defined in the disclosure are all reiterative and/or differential
and there would be considerable difficulty in applying another
embodiment instead of a computer. State of the art technology
dictates that real time calculations are biased towards comput-
ers and since there is no indication to the contrary in the
application to suggest other computational modes, this implies
the invent ion (if any) lies not in the apparatus but rather
in computer programming, the latter constituting unpatentable
subject matter.
Since the disclosed flow charting and computations are not
carried out with specific new fully disclosed apparatus de-
vised to implement a new method of boiler steam control as
a function of fuel costs, the disclosure and claims are
further rejected as being directed essentially to an algorithm
for controlling a turbine.
The applicant implies that the measurement of the primary
variables is unique and after computation of the efficiencies,
physical changes occur. It is obvious that the feedback
network which the applicant utilizes is the most standard and
well known in hydraulic flow control (United States Patent
3,676,066, Figure 1), ie, flow transmitters and variable
valve actuators. An electrical analogue is the combination
of an ammeter with a potentiometer to control current flow.
Actuation of a valve in response to a value derived from a
computer cannot be considered novel only because the computer
derives a result in a more optimum manner. The embodiments
exemplified in Figures 5, 6 and 7 show the novelty lies
solely in the flow chart or computations which are realized
by programming.
It is conclusive to the examiner that the steps recited in
claim 1 and correspondingly claims 2, 3 and 4 (same scope as
claim 1) all relate to a computer program function derived
from algorithms and flow charts and since the only embodiment
disclosed is the computer program generated one, it is
concluded the only difference from the cited structure is
the software applied.
...
In response to the Final Action the applicant stated (in part):
...
The present invention is directed to a dynamic method of
testing the units of a power plant to achieve the lowest cost
operation. Specifically, the incremental cost is equal to
the incremental efficiency (which is the ratio of incremental
input energy to incremental output energy) multiplied by the
cost of input energy. This is constantly calculated and,
as stated in the disclosure, it must be constantly calculated
since conditions change from hour to hour in most power plants
of the type considered. Thus, at any one point in time
the units with the highest and lowest incremental cost are
known. This is believed to be a significant advance in the
art.
Claim 1 defines a combination of process steps. Some of the
steps may be known in themselves, but the overall claimed
combination of process steps is submitted to be new and
unobvious. Furthermore, as actual physical process steps
are included in the claim, it is submitted to be improper to
characterize it as merely being directed to an algorithm,
computer program, or the like. Claim 1 is directed to optimiz-
ing the performance of a multi-unit power plant which produces
steam energy from a plurality of different fuels. To optimize
the performance, one step involves performing "bump" tests
and, as explained in the disclosure, this involves varying
the fuel supplies to the boilers and measuring the resultant
change in steam production, these being real, physical steps.
Claim 1 specifically recites these physical steps. Claim 1
does include, as one element of the claim, a step comprising
determining the incremental index of performance of each unit
by utilizing the incremental efficiency determination and the
cost of the fuels, but it is submitted to be incorrect to discard
the entire claim because of the inclusion in it of this step
which involves mathematical calculations. Note that the
particular calculations are not being claimed per se. Finally,
claim 1 requires the real, physical step of reallocating the energy
outputs of the units by changing the fuel inputs in accordance
with calculated indices of performance.
...
The consideration before the Board is whether or not the application is direct-
ed to patentable subject matter. Claim 1 reads as follows:
A method of optimizing the performance of a multi-unit power
plant which produces steam energy from a plurality of different
fuels where said steam energy used for independent purposes is
a significant fraction of the total energy produced by said
power plant and where input perturbations to the system affect-
ing its performance cannot be directly and accurately measured
said method comprising the following steps: determining by
bump tests the real time incremental efficiency of the units
of said power plant including the step of sensing a change in
fuel input; determining the incremental index of performance of
each of said units by utilizing said incremental efficiency
determination and the costs of said fuels; and reallocating the
energy outputs of said units by changing the fuel inputs in
accordance with said indices of performance.
As we have stated earlier this application is for optimizing the performance
of a multi-unit power plant in which the steam is used both to generate
electricity and as part of the processing operation. On page 4 of the disclos-
ure the application states at line 4 ff. that:
The steam energy used for different purposes is a
significant fraction of the total energy produced by
the power plant. Input perturbations to the system
affecting its performance cannot be directly and
accurately measured. The real time incremental effic-
iency of the units of the power plant are determined
by bump tests. Change in the energy output of the units
in response to change in energy input is sensed. The
index of performance of the units is determined by
utilizing the incremental efficiency determination.
The energy outputs of the units are reallocated in
accordance with the index of performance.
The Final Action analyzes claim 1 as having a direct correlation to
figures 5, 6 & 7 which show flow charts for understanding the specific
features of the invention. It says that the first step of claim 1 is
attained by the control computer (fig. 5), the second step is accomplish-
ed by the use of equations and the corresponding flow chart (fig. 6)
while the third step calls for reallocating the energy outputs as de-
pitted in figures 5 and 7. From this the Examiner concludes that the
"flow charts used to develop steps 1, 2 and 3 of claim 1 clearly indicate
the presence and exigency for a computer to achieve the integrated
result of claim 1".
Applicant's claim 1 specifies optimizing the performance of a multi-unit
power plant which produces steam energy from a plurality of different fuels
where the input perturbations to the system affecting its performance
cannot be directly and accurately measured. Incremental efficiency is
determined by bump tests after which the fuel is reallocated for the most
efficient operation. However no art is cited to show that "bump tests"
for a multi-unit power plant are known. Further the Schlumberger
56 CPR 2d(1981)decision states that "the fact that a computer is or should
be used to implement discovery does not change the nature of that discovery."
Claim 1 specifies a bump test in combination with a computer and in
our view does describe patentable matter in compliance with Section 2 of
the Act .
We note that U.S. patent 3,676,066 was initially cited as an applied reference
in the Final Action. This patent was cited to show "it is obvious that the
feedback network which the applicant utilizes is the most standard and well
known in hydraulic flow control". We make the following observations with
respect to this patent which relates to a chemical process such as the
production of ammonia where the unreacted feed constituents are recycled.
It shows feed components that react in fixed ratios to one another and the
conversion per pass is in the order of 25 percent. This patent is also
directed to a control procedure which permits the composition of reactor feed
"to be regulated in response to an analysis of the feed introduced into the
reactor or an analysis of the recycle stream." The application before us
does not have any fixed ratio of feed component reaction, and does not recycle
unreacted feed components but is concerned with fuel supply for the most
efficient operation.
The Final Action states that "it is well known in the systems art to employ
computers for feedback control i.e. sensing flow rates" and concludes that
since the equations as defined in the disclosure are all reiterative there
would be considerable difficulty in applying another embodiment instead of
a computer. It adds that the "state of the art technology dictates that
real time calculations are biased toward computers and since there is no
indication to the contrary in the application to suggest other computational
modes, this implies the invention (if any) lies not in the apparatus but
rather in computer programming."
We think that the above example referring to sensing flow rate is an area
where input perturbations to the system are directly and accurately measured.
This is not the condition described in this application where input perturba-
tions to the system affecting its performance cannot be directly and accurately
measured. Therefore, in the absence of cited art we do not agree that the
invention resides only in computer programming.
In summary, we recommend that refusal of the application and claims be with-
drawn, and the application be returned to the examiner. We note in return-
ing the application to the examiner that claims 2 and 4 are identical.
M.G. Brown S.D. Kot
Acting Chairman Member
Patent Appeal Board
I have reviewed the prosecution of this application and concur with the
reasoning and findings of the Board. Accordingly, I am returning this
application to the Examiner.
J.H.A. Gari‚py
Commissioner of Patents
Dated at Hull, Quebec
this 24th. day of November, 1983
Agent for Applicant
Smart & Biggar
Box 2999, Station D,
Ottawa, Ontario