Effect of Fluctuations in the Voltage Supplying Gas-Discharge Lamps on Visual Performance
Abstract Generated abstract
This paper examines how short-term voltage fluctuations in networks supplying gas-discharge, chiefly fluorescent, lighting affect visual performance, work productivity, and perceived flicker. The authors conducted three experiments with controlled fluctuation depth and frequency, measuring physiological indices of visual function, proofreading productivity and quality, subjective flicker thresholds, and comparisons with incandescent lighting. The results indicate that, within the investigated frequency range, fluctuations up to about 10 percent did not produce statistically significant impairment, whereas greater depths reduced visual and work-performance indicators. The study argues that permissible voltage deviations and fluctuations for gas-discharge lamps should differ from those for incandescent lamps and that existing standards require refinement based on further investigation.
Full Text
EFFECT OF VOLTAGE FLUCTUATIONS SUPPLYING GAS-DISCHARGE LAMPS ON VISUAL PERFORMANCE
V. D. Nikitin, L. A. Zaspanov, N. A. Dulzon
(Presented by the scientific seminar of the departments of lighting engineering and light sources
and of electrical systems and networks)
Changes in the voltage in a network over time lead to corresponding changes in the luminous flux of light sources and, consequently, in illumination. Light sources are intended to operate at the nominal voltage specified for them. At voltages differing from the nominal value, the characteristics of the sources and of the control gear change. The effect of voltage changes on the service life of lamps and control gear, as well as on the luminous flux of lamps, is especially noticeable.
The effect of voltage changes on the service life of gas-discharge lamps and control gear is the subject of special studies and is not considered in the present work.
In addition to prolonged voltage deviations in lighting networks, short-term rapid deviations often occur—the so-called voltage fluctuations—most often caused by the starting or operating mode of sufficiently powerful power consumers supplied from transformers common with the light source.
The PUE, mainly on the basis of investigations by G. M. Knorring [1], establishes the following requirements: for lighting lamps, voltage fluctuations up to 1.5% are permitted without limitation of frequency; fluctuations of 1.5–4% must occur no more than 10 times per hour; and fluctuations of more than 4%—no more than once per hour. Essentially the same requirements were included in the GOST standard for electric-power quality (GOST 13109-67), which permits voltage fluctuations at the terminals of lighting lamps (in excess of the permissible voltage deviations) depending on the frequency of their recurrence, according to the formula
\[ V_t = 1 + \frac{6}{n}, \]
where \(V_t\) is the voltage fluctuation, %,
\(n\) is the number of repetitions, h\(^{-1}\).
The same GOST permits, for lighting installations, deviations of voltage from the nominal value within the limits from minus 2.5 to plus 5%.
However, G. M. Knorring’s investigations were carried out in the prewar years on incandescent lamps.
In the literature available to us there is no information on studies similar to those of G. M. Knorring but carried out on gas-discharge lamps. At present these lamps are the most widely used, and according to SNiP II-A. 9-71 they are becoming the principal source of li—
when designing lighting installations. This makes it important to study the effect of voltage fluctuations on lighting conditions with gas-discharge lamps.
The aim of the work was an experimental study of the influence of the depth and frequency of voltage fluctuations on lighting conditions with gas-discharge lamps. Three experiments were carried out. In the first
Fig. 1. Dependence of the luminous flux of incandescent lamps and fluorescent lamps in a two-lamp circuit on voltage. 1—incandescent lamp; 2—fluorescent lamps in a two-lamp compensated circuit (total flux); 3—lagging-current lamp; 4—leading-current lamp
experiment, in order to study the effect of mains-voltage fluctuations on visual and general performance, the levels and dynamics of functions characterizing the state of the visual analyzer and of the central nervous system were investigated. As a criterion for assessing the degree of fatigue, the relative change in the levels of the studied indicators under the influence of a dosed visual load was used. Trained observers with normal vision, aged 20–25 years (4–6 persons in each series of observations), took part in the experiment.
The lighting installation was mounted in a room of nearly square shape \((9.5 \times 8.5\ \mathrm{m})\), with a height of \(3.5\ \mathrm{m}\), and with reflection coefficients \(\rho_{\mathrm{p}} = 0.7\), \(\rho_{\mathrm{st}} = 0.6\), \(\rho_{\mathrm{r}} = 0.3\). The installation consisted of 21 ShOD-2×40 luminaires with LDTs-40 lamps; the specific power was \(20.7\ \mathrm{W/m^2}\), and the illuminance was \(300\ \mathrm{lx}\).
To create periodic voltage fluctuations on the lamp–PRA set, a single-phase transformer \(T\) (Fig. 2) with a power of \(2.5\ \mathrm{kVA}\), \(U_1 = 220\ \mathrm{V}\), \(U_{2\min} = 180\ \mathrm{V}\), \(U_{2\max} = 260\ \mathrm{V}\), and dc magnetization was used. Such transformers provide a wide range of voltage regulation and sufficient speed of response [4, 5].
The control winding of transformer \(T\) is supplied from the 220-V network through a LATR autotransformer and a diode rectifier bridge. An additional resistance \(R\) is inserted into the winding circuit,
shunted by the operating contact K of the relay control circuit for voltage fluctuations. The latter, acting on contact K, makes it possible to regulate the duration and frequency of changes in the voltage-fluctuation pulses in the secondary winding of the transformer. The depth of the fluctuations is regulated by changing the resistance \(R\).
The procedure of the study itself was as follows. In the subjects, after a 15-minute adaptation period, the following were studied by standard methods: discrimination time, visual acuity, stability of clear vision, the speed and quality of performing proofreading tests, letter tests, and Landolt rings.
Fig. 2. Diagram of the test setup: 1 — control winding; 2 — lighting load; 3 — circuit of the relay control system for voltage fluctuations
Then, for 2 hours, the subjects performed strenuous visual work characteristic of electric-lamp production—rejecting filaments or electrodes. During this time the operator created (or did not create) fluctuations of a specified depth (0; 5; 10; 15%) and frequency. The subjects did not know the schedule and the nature of the voltage changes. After the work period was over, the study using the same physiological indices was repeated, and conclusions were drawn about the quality and productivity of labor over the elapsed two hours. These studies were carried out over 56 working days. The results of the studies were subjected to statistical processing.
Table 1 gives the dependence of the stability of clear vision and of discrimination time (in comparison with the initial level) on the amplitude of fluctuations for frequencies of 30 and 60 fluctuations per hour. At a significance level of \(\rho = 0.05\), the differences in the indices in the presence of voltage fluctuations and in their absence become statistically reliable at \(\Delta U \geq 10\%\).
Table 1
Stability of clear vision and discrimination time at different amplitudes and frequencies of fluctuations, and reliability of differences in the indices (\(\rho = 0.05\))
| Frequency of fluctuations, h\(^{-1}\) | Stability of clear vision: index, relative units, fluctuation depth 0% | Stability of clear vision: index, relative units, fluctuation depth 5% | Stability of clear vision: index, relative units, fluctuation depth 10% | Stability of clear vision: index, relative units, fluctuation depth 15% | Stability of clear vision: probability, fluctuation depth 5% | Stability of clear vision: probability, fluctuation depth 10% | Stability of clear vision: probability, fluctuation depth 15% | Discrimination time: index, %, fluctuation depth 0% | Discrimination time: index, %, fluctuation depth 5% | Discrimination time: index, %, fluctuation depth 10% | Discrimination time: index, %, fluctuation depth 15% | Discrimination time: probability, fluctuation depth 5% | Discrimination time: probability, fluctuation depth 10% | Discrimination time: probability, fluctuation depth 15% |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 30 | 0.9 | 0.89 | 0.85 | 0.80 | \(>p\) | \(<p\) | \(<p\) | 114 | 117 | 122 | 134 | \(>p\) | \(<p\) | \(<p\) |
| 60 | 0.9 | 0.88 | 0.82 | 0.77 | \(>p\) | \(<p\) | \(<p\) | 114 | 122 | 127 | 145 | \(<p\) | \(<p\) | \(<p\) |
The dependence of labor productivity when processing proof sheets (in comparison with the initial level) on the amplitude and frequency of voltage fluctuations, and data on the reliability of differences in the above-mentioned indicators under conditions with and without voltage fluctuations, are given in Table 2.
Table 2
Labor productivity at different amplitudes and frequencies of fluctuations and the reliability of differences in the indicators (significance level \(p=0.05\))
| Frequency of fluctuations, \(h^{-1}\) | Processing of letter proof sheets: indicator, %, amplitude 0 | Processing of letter proof sheets: indicator, %, amplitude 5 | Processing of letter proof sheets: indicator, %, amplitude 10 | Processing of letter proof sheets: indicator, %, amplitude 15 | Processing of letter proof sheets: probability, amplitude 5 | Processing of letter proof sheets: probability, amplitude 10 | Processing of letter proof sheets: probability, amplitude 15 | Processing of Landolt-ring tables: indicator, %, amplitude 0 | Processing of Landolt-ring tables: indicator, %, amplitude 5 | Processing of Landolt-ring tables: indicator, %, amplitude 10 | Processing of Landolt-ring tables: indicator, %, amplitude 15 | Processing of Landolt-ring tables: probability, amplitude 5 | Processing of Landolt-ring tables: probability, amplitude 10 | Processing of Landolt-ring tables: probability, amplitude 15 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 30 | 97 | 96 | 93 | 90 | \(>p\) | \(<p\) | \(<p\) | 100 | 99 | 97 | 96 | \(>p\) | \(<p\) | \(<p\) |
| 60 | 97 | 95 | 91 | 88 | \(<p\) | \(<p\) | \(<p\) | 100 | 99 | 96 | 94 | \(>p\) | \(<p\) | \(<p\) |
The dependence of the quality indicators for the processing of proof sheets (in comparison with the initial level) on the amplitude of fluctuations for frequencies of 30 and 60 fluctuations per hour, and data on the reliability of the differences, are contained in Table 3.
Table 3
Quality of processing proof sheets at different amplitudes and frequencies of fluctuations and the reliability of differences in the indicators (significance level \(p=0.05\))
| Frequency of fluctuations, \(h^{-1}\) | Quality of processing letter proof sheets: indicator, %, amplitude 0 | Quality of processing letter proof sheets: indicator, %, amplitude 5 | Quality of processing letter proof sheets: indicator, %, amplitude 10 | Quality of processing letter proof sheets: indicator, %, amplitude 15 | Quality of processing letter proof sheets: probability, amplitude 5 | Quality of processing letter proof sheets: probability, amplitude 10 | Quality of processing letter proof sheets: probability, amplitude 15 | Quality of processing Landolt-ring tables: indicator, %, amplitude 0 | Quality of processing Landolt-ring tables: indicator, %, amplitude 5 | Quality of processing Landolt-ring tables: indicator, %, amplitude 10 | Quality of processing Landolt-ring tables: indicator, %, amplitude 15 | Quality of processing Landolt-ring tables: probability, amplitude 5 | Quality of processing Landolt-ring tables: probability, amplitude 10 | Quality of processing Landolt-ring tables: probability, amplitude 15 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 30 | 100 | 99 | 95 | 93 | \(>p\) | \(<p\) | \(<p\) | 100 | 99.5 | 99 | 98.5 | \(>p\) | \(<p\) | \(<p\) |
| 60 | 100 | 97 | 94 | 92 | \(>p\) | \(<p\) | \(<p\) | 100 | 99.5 | 99 | 98 | \(<p\) | \(<p\) | \(<p\) |
In the second experiment, carried out on the same installation, the subjective evaluation of voltage fluctuations in a network with gas-discharge lamps was investigated in order to establish the curve of threshold visibility of flicker and the curve of the irritation threshold. The experiment consisted in the amplitude of the voltage fluctuations being gradually increased from zero. The threshold value was taken to be that at which 75% of the observers detected a change in illumination. During the experiment, 6 subjects performed visual work of medium accuracy (class IVg according to SNiP-71) and, on special forms, recorded the time when they noticed fluctuations in illumination (brightness). The study of the thresholds of flicker visibility showed (Fig. 3) that the corresponding values of voltage fluctuations for incandescent lamps—
lighting [6] and fluorescent lamps (in the region we investigated) differ substantially.
The purpose of the third experiment was to compare the effect of voltage fluctuations on illumination by gas-discharge lamps and incandescent lamps. The experiment was carried out in a chamber \(1.4 \times 1.4 \times 1.4\ \text{m}\) with white walls \((\rho_{\text{st}} = 0.8)\) and a luminous ceiling, illuminated by the translucent luminaire OD-2×40 with LB-40 lamps, or by 6 incandescent lamps. The brightness of the ceiling was the same both with gas-discharge lamps and with incandescent lamps; the illuminance of the working surface in the chamber reached 360 lx. To produce periodic voltage fluctuations, the same regulator was used (Fig. 2).
Fig. 3. Effect of voltage fluctuations on the threshold of flicker visibility:
1 — threshold of flicker visibility with incandescent lamps; 2 — threshold of flicker visibility with fluorescent lamps.
In this experiment, we investigated the dependence of productivity and quality of work in processing proofreading samples on fluctuations of the voltage supplying either incandescent lamps or fluorescent lamps. The results are presented in Fig. 4 and show a decrease in the number of characters viewed and a reduction in the quality of sample execution with increasing amplitude and frequency of fluctuations, as well as in the transition from fluorescent lamps to incandescent lamps.
Fig. 4. Dependence of productivity (a) and quality (b) of work on voltage fluctuations in the network with gas-discharge lamps and incandescent lamps: — fluorescent lamps; —·— incandescent lamps; 1, 4, 7, 10 — 10 fluctuations/h; 2, 5, 8, 11 — 30 fluctuations/h; 3, 6, 9, 12 — 60 fluctuations/h.
Let us briefly consider voltage deviations. It is known that, in fluorescent lamps, when there are voltage deviations in the network, the luminous
the flux changes less than in incandescent lamps [2, 3]. From Fig. 1 it is not difficult to see that the same relative decrease in luminous flux for fluorescent lamps with a two-lamp compensated supply circuit, as compared with incandescent lamps, is observed at voltage deviations 2–3 times greater, i.e., for fluorescent lamps substantially larger voltage deviations are permissible than for incandescent lamps.
Taking into account that the change in luminous flux under voltage deviations also depends on the particular types of light sources and on the circuits by which they are connected to the network, additional studies are necessary in order to resolve the question of permissible voltage deviations in a network supplying fluorescent lamps.
The authors express their gratitude to N. I. Beloborodova, V. A. Kuryanovich, and G. G. Ermolov for their participation in setting up and conducting the experiments.
Conclusions
-
Voltage fluctuations in a network with gas-discharge lighting lamps, in the investigated range of frequencies and at a depth of up to 10%, do not affect the functions characterizing the state of the visual analyzer and of the central nervous system. At a fluctuation depth of more than 10%, a statistically significant decrease was established in visual and other functions, as well as in quantitative and qualitative indices of work.
-
In networks with gas-discharge lamps, larger voltage deviations are permissible than in networks with incandescent lamps.
-
The GOST standard for the quality of electric power and the Electrical Installation Code require clarification of the permissible deviations and fluctuations of voltage for gas-discharge lamps; for this purpose, additional studies must be carried out.
LITERATURE
-
G. M. Knorring. The significance of voltage quality for lighting installations. In: Reports at the All-Union Conference on Voltage Quality and Its Regulation in Electrical Networks and Systems. Moscow, 1961.
-
N. V. Volotskoy et al. Electric Lighting of Industrial and Civil Buildings. Moscow–Leningrad, “Energiya,” 1964.
-
A. P. Ivanov. Electric Light Sources. Moscow, Gosenergoizdat, 1955.
-
A. M. Bamdas, S. V. Shapiro. Transformers Controlled by Biasing. Moscow, “Energiya,” 1965.
-
A. M. Bamdas, V. A. Somov, A. O. Shmidt. Transformers and Stabilizers Controlled by Biasing. Moscow, Gosenergoizdat, 1959.
-
El. Power Distribution for Ind. Plants, 1956. Cited in N. V. Volotskoy et al., Electric Lighting of Industrial and Civil Buildings. Moscow–Leningrad, “Energiya,” 1964.