Design of a FPGA-based controller for power and period measurement in the start range of Dalat Nuclear Research Reactor

Nuclear Science and Technology, Vol.10, No. 3 (2020), pp. 41-48  
Design of a FPGA-based controller for power and period  
measurement in the start range of Dalat Nuclear Research Reactor  
Vo Van Tai, Nguyen Van Kien, Nguyen Nhi Dien, Trinh Dinh Hai, Le Van Diep  
Dalat Nuclear Research Institute  
01 Nguyen Tu Luc, Dalat city, Vietnam  
(Received 03 August 2020, accepted 10 November 2020)  
Abstract: This paper introduces a new controller module based on a high-speed field-programmable  
gate array (FPGA) and digital signal processing (DSP) using moving average (MA) filters for  
calculation of the reactor power and period at the start range of the Dalat nuclear research reactor  
(DNRR). The reactor power is proportional to the neutron flux in the reactor core, and the reactor  
period is the time that the reactor power changes by a factor of 2.718. In the control and protection  
system (CPS) of the DNRR, the reactor power and period have been monitored by the 8-bit  
microprocessor controller named BPM-107R. There are two main functions of the BPM-107R  
controller including 1) measurement and determination of reactor power and period and 2) generation  
of warning and emergency protection signals by reactor power or/and by reactor period. Those  
discrete signals will access to the logical processing unit of the CPS to prohibit the upward movement  
of control rods or to shut down the reactor. The CPS has three BPM-107R controllers corresponding  
to three independent neutron flux measurement equipment (NFME) channels working by logic voting  
“2 out-of 3”. Each NFME channel was designed for detection of neutron flux density in the full range  
from 1×100 to 1.2×1010 n/cm2×s, which is divided into two sub-ranges named start range (SR) and  
working range (WR). The designed FPGA-based controller module was tested using simulated signals  
as well as signals from the CPS in comparison with the original controller BPM-107R. The  
experimental results show that the characteristics and functions of the two controllers are equivalent.  
Keywords: Reactor period, reactor power, moving average filter, FPGA, DSP.  
(Ppreset can be set from 1% to 100% of 500  
kWt by 1% step) or the period is less than 20  
s, the emergency protection signal EMR-P or  
EMR-T will be generated to shut down the  
reactor and hold it at a subcritical state by  
inserting neutron-absorbing control rods into  
the reactor core. The above-mentioned  
functions have been realized by the existing  
embedded micro-controller BPM-107R using  
DS87C530 processor. Main characteristics of  
this controller include the working clock  
frequency is of 30 MHz, the data memory  
(RAM) is of 32 kbyte, the programme  
memory (ROM) is of 64 kbyte and the length  
of each neutron parameter generation cycle is  
of 200 ms [1, 2].  
I. INTRODUCTION  
The reactor power (P) and reactor period  
(T) are parameters directly related to safety  
operation of a nuclear reactor. In reactor  
operation condition, the time rate of power  
change is very important, particularly at the  
start-up stage, then monitoring the reactor  
period is an effective way to prevent the  
reactor power from increasing rapidly.  
Therefore, the on-line and accuracy  
calculations of reactor power and period have  
been required to avoid uncertain shutdown or  
nuclear incidents.  
In case of the CPS of DNRR, when the  
reactor power is greater than Ppreset by 10%  
©2020 Vietnam Atomic Energy Society and Vietnam Atomic Energy Institute  
DESIGN OF A FPGA-BASED CONTROLLER FOR POWER AND PERIOD MEASUREMENT...  
The digital CPS of the DNRR was  
designed and supplied by the SNIIP  
SYSTEMATOM JSC of Russia and put into  
operation in 2007. It was the second generation  
of the instrumentation and control (I&C) system  
for nuclear reactors that was designed by this  
JSC, meanwhile, the fifth generation of I&C  
system has been operated at some nuclear  
reactors since 2015. As usual, after 13 years of  
operation, many functional electronic modules of  
the CPS were out of products in the market for  
spare parts and replacement. Besides, taking into  
accounts on the advantage of the FPGA  
technology for design and construction of nuclear  
electronic equipment, nowadays fast integrated  
FPGA-based devices are widely developed for  
measurement, data storage and acquisition  
system in nuclear physics experiments and  
nuclear engineering, such as for radiation  
measurements with high resolution X-ray  
spectroscopy with pulse height analyzer [3], for  
readout systems of the cosmic ray angular  
distribution based on flash analog-to-digital  
converter using FPGA technology [4], and for  
nuclear data measurement by (n, γ) and (n, 2γ)  
reactions at horizontal experimental neutron  
beam ports of nuclear research reactor with high  
performance  
gamma  
spectrometers  
[5].  
Considering the above-mentioned reasons and  
propensity, the main purpose of this work is to  
develop the new FPGA-based controller module  
which can be a spare-part or replacement for the  
BPM-107R controller when necessary. The  
developed controller applies DSP principle with  
moving average filtering technique using a Xilinx  
FPGA Artix-7 with on-board clock of 50 MHz  
and two 32-bit counting channels that can sample  
and calculate reactor power pulse signals every  
20 ms. Some comparisons with the current BPM-  
107R controller module were performed by  
simulated signals as well as by real signals from  
neutron detectors of the start range.  
S1  
S2  
S3  
C1
1
C2  
C3  
L1  
R1  
L2  
L3  
R3  
D D  
R2  
A
Fig. 1. The "2 out-of 3" data processing and architecture of the DNRR’ CPS  
(S – sensor/detector, C – Controller, L – Logic processing, R – Relay block, A – Actuator,  
D – Display panel)  
The CPS of DNRR is operating based on  
neutron detectors are installed in dry channels  
outside the reactor core, as shown in Fig. 2.  
“2 out-of 3” principle as shown in Fig. 1. Three  
42  
VO VAN TAI et al.  
Fig. 2. Neutron detector unit BDPN-36R2 (left) and detector locations in reactor pool (right)  
Where, P(t) is transient reactor power at  
time t, P0 is initial reactor power at time t0, T is  
the time taken for the power to change by a  
factor e (e = 2.718 which is the base of natural  
logarithm) that called the reactor period, t is  
time during the reactor transient.  
II. METHODS  
A. Nuclear reactor power and period caculation  
Neutron density (power) of a nuclear  
reactor follows the exponential function of:  
 
 
           
(1)  
 
Fig. 3. Sampling for nuclear reactor power and period calculation  
To calculate the reactor period from (1),  
it is necessary to convert continuous time  
models as Fig. 3 to discrete time form, using  
finite difference techniques. The reactor period  
in digital nuclear instrumentation system can  
be calculated by the equation (2) [6]:  
From equation (2) we can see that the  
reactor period can reflect the state of a  
nuclear reactor. When Pk-Pk-1 = 0, the reactor  
is in a stable state (T〜999 s); when Pk-Pk-1  
>
0, the reactor power increases; and when Pk-  
Pk-1 < 0, the reactor power decreases. The  
reactor period can be monitored from 0 to  
999 s by 1 s step.  
 
 
   
     
 
 
    
    
or  
 
 
 
or  
   
   
 
 
    
 
 
   
 
 
The reactor power level is proportional  
to the reactor neutron flux, therefore, the  
output frequency from a pulse amplifier is  
proportional to the reactor power level, then  
the relationship between the reactor power at  
 
 
    
     
(2)  
     
 
    
Where, Pk-1, Pk is the (k-1)-th and k-th  
sampled reactor power, △t is the sampling  
period in seconds.  
43  
DESIGN OF A FPGA-BASED CONTROLLER FOR POWER AND PERIOD MEASUREMENT...  
the start range (PSR) and the output frequency  
of the pulse amplifier can be calculated as:  
inputs and makes an output with them, the  
formula is expressed as follows: [8, 9]  
PSR = KSR×FSR×10-6  
(3)  
n1  
P PM 1 .....P  
1
M
M (n1)  
P   
P
M i  
n   
i0  
SM  
n
Where, PSR is the reactor power at the  
start range, KSR is a coefficient, FSR is the  
output frequency from the pulse amplifier  
which is connected to the fission chamber for  
monitoring in the range from 10-6 to 10-1 %  
Pnominal (Pnominal = 500 kWt).  
(4)  
When calculating sequential values, a  
new value comes into the sum, and the oldest  
value moves out, n is coefficient or window of  
filter. Equation (4) can be rewritten as:  
B. Design of moving average filters using  
FPGA and DSP  
1
P PSM , pre (P P  
)
(5)  
SM  
M
M n  
n
A moving average (MA) filter is a  
type of finite impulse response filter (FIR)  
[7]. It is used widely in many applications  
P
Where, SM is the average value,  
P
SM , pre  
is the previous average,  
P is the new sample,  
M
in  
digital  
signal  
processing,  
P
is the n-th old sample, n is the window  
communications, control, electrical and  
biomedical systems. The main purpose of  
MA filter is to increase the signal to noise  
ratio and to reduce random noise.  
M n  
of SMA filter.  
Equation (5) can be described in Fig. 4,  
in which the reactor power and the reactor  
period are calculated by DSP technique  
embedded on FPGA.  
A simple moving average (SMA) filter  
is a filter that averages N points of previous  
SMA Filter  
Xi  
Pulse  
circuit  
Yi  
SUB  
DIV  
ADD  
SAMPLE IN  
Clock  
SAMPLE OUT  
Z(-n)  
Z(-1)  
n
Fig. 4. Principle schema used for moving average filter  
for calculating reactor period  
T
1
2
3
F_T  
1
2
EMR_T  
MULT  
MULT  
3
SUB  
INVERSE  
Comp5  
Comp6  
set_Scram T=20s  
Clock  
Z(-1)  
WRM_T  
Dt  
set_Alarm T=40 s  
Fig. 5. Principle schema used for calculating reactor period  
44  
VO VAN TAI et al.  
SAMPLE IN is the input of the SMA  
then calculated period is compared with  
set_Alarm T (40 sec) and set_Scram T (20 sec)  
to generate discrete signals WRN_T and  
EMR_T, respectively. Further, these signals  
will access to the logical processing unit of the  
CPS, in which signal WRN_T is used for  
formulation of a command to prohibit the  
upward movement of control rods, meanwhile  
EMR_T signal formulates a command to shut  
down the reactor.  
filter that is processed by a pulse circuit for  
normalizing the pulse width, then counting,  
accumulating and performing SMA filtering  
technique, the output signal (SAMPLE  
OUT) is frequency, n is a coefficient of  
filter that is dependent on the fluctuation of  
input signal.  
The diagram for reactor period  
calculation by Eq. (2) can be described in Fig. 5.  
To calculate the reactor power at the start  
range, Eq. (3) is described in Fig. 6.  
F_T is the output signal from SMA filter  
that is used for calculating reactor period, and  
Psr  
for calculating reactor power Psr  
F_Psr  
1
3
MULT  
1
2
EMR_Psr  
DIV  
3
2
Comp1  
Comp2  
Clock  
set_Scram P  
Ksr  
1000000 set_Alarm P  
WRN_Psr  
Fig. 6. Principle schema used for calculating reactor power at start range  
As shown in Fig. 6, the F_Psr signals are  
controller module in the start range (named  
FPGA-SR). The experiment which was setup  
to measure the power and the period for  
testing FPGA-SR and BPM-107R controllers  
is shown in Fig. 7(a). The values of power  
and period of both the controllers were  
recorded (data logging) using a computer via  
Terminal v1.9b software.  
performed to calculate the reactor power for the  
start range (Psr). Then these values are  
compared with set_Alarm P (5% Ppreset) and  
set_Scram P (10% Ppreset) for generating discrete  
signals WRN_Psr and EMR_Psr, respectively.  
Further, these discrete signals will access to the  
logical processing unit of the CPS, in which  
signal WRN_Psr is used for formulation of a  
command to prohibit the upward movement of  
control rods, meanwhile EMR_Psr signal  
formulates a command to shut down the reactor.  
In the CPS, the KSR coefficient for  
calculating reactor power was calibrated to be  
4.04. The initial value of simulated frequency  
was 10 Hz and the final value was 50,000 Hz.  
The simulated period was chosen to be 20 s.  
The relative error of the monitored reactor  
power in the range from 4×10-5 to 2×10-1 %  
Pnominal is within 5% and that of the monitored  
reactor period is within 10%, which are  
reported in [10]. Figure 7(b) shows the testing  
results for period measurement at 20 s by the  
BPM-107R and FPGA-SR controllers. The  
III. EXPERIMENTAL RESULTS  
AND DISCUSSION  
A. Testing of the designed controller in the  
start range by simulator  
The simulator module PGT-17R,  
designed by SNIIP SYSTEMATOM JSC,  
was used for testing of the new designed  
45  
DESIGN OF A FPGA-BASED CONTROLLER FOR POWER AND PERIOD MEASUREMENT...  
period curve of the BPM-107R controller  
reaches to 20 s slower than that of the FPGA-  
SR controller, in which the period curve  
reaches to 20 s only after about 100 s.  
BPM-107R  
F
COMPUTER  
COMPUTER  
Data logging  
LCD&KEYPAD  
PGT-17R MODULE  
FPGA-SR  
KEYPAD  
F: 10 to 50 000 Hz  
Ksr=4.04  
LCD&KEYPAD  
(a)  
(b)  
Fig. 7. Principle schema for measurement of power and period using PGT-17R simulator module (a)  
and Testing results of BPM-107R and FPGA-SR controllers using PGT-17R module with FSR from 10 Hz to  
50,000 Hz at 20 s period (b)  
The main reason of different results  
between two controllers is due to their different  
architectures. The BPM-107R controller is  
based on the 8-bit DS87C530 microprocessor  
operating with 30 MHz clock frequency that  
can sample and calculate reactor power pulse  
signals between 20 ms to 200 ms. Meanwhile,  
the new developed controller is based on the  
high speed FPGA with 50 MHz on-board clock  
and two 32-bit counting channels that can  
sample and determine the above procedures  
every 20 ms. The other reason is that in the  
range of reactor power lower than 5×10-4 %  
Pnominal, the statistic counting of simulated  
pulses is still very low, therefore, in order to  
achieve the ideal curve from 5×10-4 % Pnominal  
(from this value the emergency protection is  
started monitoring), the filter coefficient in the  
BPM-107R was changed as jumping steps.  
Instead of that, the coefficient n (or window) of  
SMA filters in the FPGA-SR controller with  
embedded DSP circuits can be flexibly  
changed to achieve the such ideal curve.  
Fig. 7(b) also shows that both the BPM-  
107R and FPGA-SR controllers give almost  
the same results in the case of simulation for  
power measurement.  
B. Testing of the designed controller by  
neutron signals in start range of the reactor  
The experimental schema to measure the  
reactor power and period at the CPS using  
neutron signals from neutron detectors in the  
reactor is shown in Fig. 8.  
FSr  
FSr  
BPM-107R  
SIGNAL  
F_OUT  
+400VDC  
PULSE  
LCD&KEYPAD  
COMPUTER  
AMPLIFIER  
-400VDC  
Data logging  
FPGA-SR  
+12VDC  
-12VDC  
LCD&KEYPAD  
Fig. 8. Principle schema for measurement of reactor power and period  
46  
VO VAN TAI et al.  
Fig. 9. Results of reactor power and period measurement by FPGA-SR and BPM-107R controllers in  
the range from 4×10-5 to 2×10-2 % Pnominal  
Fig. 10. Results of reactor power and period measurement by FPGA-SR and BPM-107R controllers in  
the range from 10-2 to 2×10-1 % Pnominal  
The start range was designed for  
monitoring the reactor power in the range from  
4×10-6 to 2×10-1% Pnominal. In practice, the  
lowest power level is about 4×10-5 % Pnominal  
due to epithermal and fast photoneutrons,  
filters can be flexibly changed, so its response  
time is faster than that of the BPM-107R  
controller as shown in Fig. 7(b).  
IV. CONCLUSIONS  
9
4
which are produced by the Be(γ, n)2 He and  
Based on the application of DSP  
technique with moving average filters  
embedded on FPGA, the functional circuits for  
the controller module have been designed for  
measuring and determining the reactor power,  
the reactor period, and generating the  
protection signals whenever the value of the  
reactor power or period is beyond the preset  
threshold value.  
4
9Be(n, 2n)2 He reactions, respectively, are  
significant in the reactor core, especially when  
the reactor operates frequently.  
Fig. 9 and Fig. 10 show the results of the  
reactor power and period measurement in the  
range from 4×10-5 to 2×10-1% Pnominal by the  
FPGA-SR and BPM-107R controllers. The  
results show that both the controllers have  
almost the same value.  
The developed FPGA-based controller  
was tested in comparison with the BPM-107R  
module of the CPS, which was designed by  
As mentioned above, in case of the  
FPGA-SR controller, the coefficient n of SMA  
47  
DESIGN OF A FPGA-BASED CONTROLLER FOR POWER AND PERIOD MEASUREMENT...  
SNIIP SYSTEMATOM, Chief Designer А.  
А. Zaikin, 2006.  
SNIIP SYSTEMATON JSC of Russia. The  
testing was performed both by pulse-generated  
simulator PGT-17R and by real neutron signals  
from the reactor. The experimental results of  
measuring reactor power and period by the  
FPGA-SR and BPM-107R controllers indicate  
a good agreement of their functions and  
characteristics, including the reactor power  
measured in the range from 4×10-5 to 2×10-1 %  
Pnominal with the accuracy less than 5% and the  
reactor period is monitored in the range from 1  
to 999 s with the accuracy less than 10%.  
Additionally, the emergency protection signal  
EMR-P or EMR-T is generated, when the  
reactor power is greater than Ppreset by 10% or  
the period is less than 20 s.  
[2]. Safety Analysis Report for the Dalat Nuclear  
Research Reactor, Dalat Nuclear Research  
Institute, Chief Editor Nguyen Nhi Dien, 2012.  
[3]. Buzzetti S., Capou M., Guazzoni C. et al.,  
“High-speed FPGA based pulse-height  
analyzer  
for  
high  
resolution  
x-ray  
spectroscopy”, IEEE Trans. Nucl. Sci. 52, pp.  
854–860, 2005.  
[4]. V. H. Hai, N. Q. Dao and M. Nomachi,  
“Cosmic ray angular distribution employing  
plastic scintillation detectors and flash-  
ADC/FPGA-based  
readout  
systems”,  
Kerntechnik, 77, pp. 462–464, 2012.  
[5]. Pham Dinh Khang, Nguyen Nhi Dien, Dang  
Lanh, Nguyen Xuan Hai, Pham Ngoc Tuan,  
Nguyen Duc Hoa, Nguyen An Son, “A design  
configuration of an FPGA-based coincident  
spectrometry system”, J. of Analytical  
Sciences, Methods and Instrumentation, 2164–  
2745, Vol. 3, pp. 158–162, 2013.  
The obtained results also permit to  
conclude that the new developed FPGA-based  
controller meets the design purposes and can  
perform all functions of the existing BPM-  
107R controller, therefore, it can replace the  
BPM-107R controller when needed.  
[6]. E. P. Gytfopoulos and P.M. Coble, A digital  
nuclear reactor control system, 1960.  
ACKNOWLEDGEMENTS  
[7]. Huasheng Xiong, Duo Li, “Nuclear reactor  
doubling time calculation using FIR filter”,  
Energy Procedia, 39, pp. 3–11, 2013.  
The authors are thankful to Dalat  
Nuclear Research Institute and Vietnam  
[8]. “Moving  
average”  
Atomic  
Energy  
Institute  
for  
their  
administrative assistance. This research is  
supported by Ministry of Science and  
Technology of Vietnam under grant number  
ĐTCB. 08/19/VNCHN.  
[9]. The Scientist and Engineer's Guide to Digital  
Signal Processing, by Steven W. Smith,  
Chapter 15 “Moving average filters” pp. 277–  
REFERENCES  
[10]. Complex of Equipment for Control and  
Protection System ASUZ-14R, Passport  
[1]. Complex of Equipment for Control and  
Protection System ASUZ-14R, Operating  
Manual RUNK.506319.004 RE-E, JSC  
RUNK.506319.004  
SYSTEMATOM, Chief  
Zaikin, 2006.  
PS-E,  
JSC  
SNIIP  
Designer А. А.  
48  
pdf 8 trang yennguyen 10440 Free
Bạn đang xem tài liệu "Design of a FPGA-based controller for power and period measurement in the start range of Dalat Nuclear Research Reactor", để tải tài liệu gốc về máy hãy click vào nút Download ở trên

File đính kèm:

  • pdfdesign_of_a_fpga_based_controller_for_power_and_period_measu.pdf