1. Introduction

Electric-to-mechanical energy conversion by dint of electric motors allows easily and economically to actuate diverse operating mechanisms such as conveyers, lifting-conveying machines, pumps, fans, compressors, metal-cutting machine tools, rolling mills, garment equipment and so on. Due to the construction simplicity, high reliability and the democratic price squirrel-cage induction motors (IMs) are considered to be universal electric motors. Over 85% of all electric machines are 3-Phase IMs. By statistics there are approximately 50 millions of the 3-Phase IMs that run on 0.4 kW in public production of Russia now. Life expectancy of IMs without a complete overhaul is usually 15-20 years providing they are put to their proper operation. A proper IM operation means an IM operates in conformance with nominal parameters given in the IM ratings. However, more often that not IMs don't operate at rated duties because of the poor supply voltage quality and the violations of standard code such as technological overloads, elevated ambient humidity and temperature, insulation resistance drop and refrigeration system failures leading to fault IM conditions. Through the faults up to 10% of all IMs used go unserviceable every year. For instance, 60% of borehole electric pump units are broken down more often than once a year. Breakdown of IMs causes serious faults and big property damage that is associated with technology process outages, fault-induced consequences elimination, IMs trouble-shooting and repair of IMs failed. Repair of an electric machine of up to 1kWt power stands in 5-6 US$. You can evaluate the cost of more powerful machine repair by yourself. Besides an operation in abnormal conditions gives rise to increased power consumption from AC mains and reactive power consumption augmentation. Self-evidently that application of reliable and effective protection against fault conditions will considerably reduce a number and frequency of faults, power consumption and operating expenses as well as extend an IM service life. But in order to find this protection one has to know not only how and what his IM must be protected from but also the particularities of the processes that run in IMs on faults.

2. Fault IM conditions

IM faults. There are two basic types in IM faults: mechanical ones and electric ones. Deformation or break-down of rotor shaft, slackening of stator core fixing to frame, slackening of rotor core compression joint, babbitt fusion in plain bearing, failure of separator, ring or ball in antifriction bearing, impeller break-down, dust and dirt deposit in moveable elements and so on are regarded as the mechanical faults which in most cases are caused by the mains phase imbalance radial vibrations, the mechanical motor shaft overloads, accessories/components or assembling reject. Up to 10% of all IM faults will be mechanical ones with 8% of them caused by the phase imbalance and only 2% of them caused by the mechanical overload. Since a ratio of the faults induced by reject is very small we will not consider it in this analysis. The mechanical fault probability has never been estimated, the main part of mechanical faults is latent and getting exposed only after appropriate tests or after IM disassemblies have been carried out. However, AC mains voltage monitoring and an IM shaft load monitoring in most cases allow to minimize this probability.
Electric IM faults, in its turn, come in three types:

IM mains faults. Electric energy quality is defined in accordance with a variety of special standard rates, mainly, with voltage and frequency distortion, even or odd harmonic factor, negative and zero voltage sequence factors and so on. Because of supply substation faults, short-circuits in power distribution systems, switching or lightning disturbances and phase load distribution irregularity the actual values of many rates exceed the allowable ones. This is the reason of the IM fault conditions. By statistics up to 80% of the IM faults are directly or indirectly induced by AC mains voltage faults.

Analysis of electric energy quality indexes against IM operating conditions shows, for instance, that AC mains voltage stepping-down causes the stator current increase which leads to the intensive heating of IM insulation and its life reduction due to the accelerated insulation ageing and the insulation disruption. AC mains voltage steeping-up, in its turn, results in a stator magnetic flux increase, the magnetization current increase, the core heating (up to steel «fire») and the reactive power consumption augmentation that decreases a power factor.

Generalized data illustrating how the basic electric energy quality rates affect IM conditions are given in the Table ¹1.

Table ¹ 1

It should be noted that there are several mains fault types that occur very frequently but are not regulated by the Standard code because they are considered to be extreme cases of IM unbalanced conditions. These are a phase loss, incorrect phase sequence and phase «coincidence».
Phase loss is usually associated with a service cable conductor breakage, a burned-out fuse,
a circuit-breaker tripping in one of the lines or a line failure. In Wye-connection of IM windings two phase voltage divides equally and makes a half of the line voltage Uphase= Uline/2, a voltage in the third phase is absent. These conditions lead to the higher mains power consumption and the stator overheating windings. The rotating field will convert in the pulsating field, there will be no current in the missing phase, currents in two other phases will increase by 50%. The IM will not turn even at no-load conditions. There are some IM types which in case of a phase loss during IM operation generate a so-called «regeneration» voltage on the missing phase that is close to mains voltage in phase and in peak value. Then the IM goes in retardation condition and not being turned off will burn-out for a few seconds.
A faulty condition associated with phase «coincidence» occurs when a missing mains phase is short-circuited by IM to the other phase. In this case the same phase voltage is applied on two phases of the IM and a voltage that is applied on the third phase remains of a normal level. A minor peak value imbalance is followed by a considerable phase imbalance that induces considerable negative sequence voltages that lead to the IM overheating and failure.
The incorrect phase sequence (B-C-A, C-A-B instead of A-B-C) initiates an IM reversible condition when the IM rotates into the opposite side which is often forbidden by technology process conditions because it can lead not only to the IM fault but also to heavy catastrophic effects.
Continuous AC mains voltage presence & quality monitoring, including harmonic analysis, RMS or average voltage value calculation before an IM cut-in, the IM operating condition monitoring and monitoring of phase voltage parameter changes induced by the IM conditions will often allow to avoid causes of the fault condition, to prevent the short-circuit & current overload conditions.
IM current faults.
A voltage at IM terminals and IM winding phase currents are in close interconnection and any, even minute, mains voltage variations induce considerable changes in the phase currents (see table ¹.1). In order to provide an IM effective protection the phase current measurement should be performed as accurately as possible. According to recent studies, a long-term IM operation under current overload that exceeds the nominal current by 5% reduces the IM life tenfold. Since the current curve shows the distinct sinusoidal distortion, especially at starts, it carries a great number of higher harmonics which exert a considerable effect on RMS current value. That’s why a decision about IM operations must be made on calculated RMS current values because if the decision is made on average signals or, what even worse, on peak values, false conclusion about no-current overload condition can be reached.
IM current overload comes in two types: balanced overload and unbalanced overload.
The balanced current overload is generally associated with the mechanical overloads on an IM shaft. Their values depend directly on the IM conditions and heat overloads that will be described below.
The major part of IM current faults is primarily associated with the abnormal conditions inside IMs that induce the unbalanced current overload. Let’s examine the basic types of these faults (see table ¹2).

Table ¹ 2. Effect of internal abnormal conditions on IM operation.

All internal IM faults are accompanied by a considerable phase current unbalance that several times exceeds the voltage imbalance. The constant current monitoring, therefore, allowing to control current imbalance – phase imbalance relation provides instant recognition of these faults and alarms users to cut out the IM fast.

IM conditions.

There are four basic nominal IM conditions that are defined by load behavior: long-term, short-term, intermittent and mixed cycles. Without going into details of the cycles mentioned above it should be noted that a basic under-load operation characteristic is IM thermal response. An IM operation is always accompanied by its heating caused by processes that occur in the IM and energy losses. Design IM life is eventually depending on its allowable insulation heating temperature. Modern IMs are classified by A, E, B, F, H insulation classes with allowable heating temperatures as follows: A – 105°C, E– 120°C, B – 130°C, F – 155°C, H – 180°C, C –over 180°C. Exceeding of allowable temperature causes premature insulation destruction and considerably reduces an IM life.

However IM operations in service generally differ from the standardized ones. More common is the quick varying load condition when the IM goes into overload condition, then returns to the design condition or the under-nominal-load condition. If the machine is operating in the long-term condition with varying load (P1, P2, P3…), an unsteady-state heat process is taking place (see fig. 4), because different power losses and, therefore, different heat losses arise in it for different time intervals t1, t2, t3, t4. In order to control effectively the heat quality accumulated by an IM in the operation one needs to find out the IM heating and cooling laws.

IM heat balance equation

Because of difficulties of carrying out the analysis the following basic assumptions of IM heat balance model are made: an IM is considered to be an uniform body with infinite thermal conduction and equal point-by-point temperature. The IM heat capacity and heat transfer factor don’t depend on IM shaft load. The IM temperature is accepted as the difference between an actual temperature value and an ambient value and refers to a relative temperature value
Temperatures of the different motor parts are proportional to each other and the motor in the whole can be characterized by a mean temperature T. It’s implied that
the heating process occurs uniformly and the different motor part temperatures are steady-state.
The mean temperature T is proportional to the quantity of heat Q accumulated in the motor.
T=Q / C (1)
where C is a heat capacity of the motor.
The heated motor thermal loss are proportional to its temperature.
dQ/dt = - k*T = - k*Q/C (2)
where k is a heat conduction between the motor and ambient air.
It’s supposed that the motor was cold before start.
Following this model the basic heat equation at the motor operation is
dQ/dt = -k*Q/C + I2R (3)
where I2R is the power released in the motor during the passage of current I through windings with resistance R.
The solution of the equation at direct current I
Q(t) = Q0 *(1-exp(-t*(k/C))) (4)
The steady–state (dQ/dt=0 ) quantity of heat in the motor is Q0= I2*R*C/k
To the maximum safe motor current I nom correspond the maximum safe quantity of heat
Q nom = I2nom*R*C/k (5)
and the maximum permissible temperature (relative to environment)
T nom = Q nom /C = I2 nom *R/k (6)
At the motor start on direct current that N times exceeding the maximum safe I nom
the maximum safe heat quantity period TN is
TN = (ln(N2)-ln(N2-1))/(k/C) (7)
Majotiry of IM heat overload protections is based on the quasithermal mathematical models of IMs. Continuous I2 calculation allowing for IM heating and cooling rate with as high as possible measurement digitizing degree gives a complete view of IM accumulated heat quality which is considered as dangerous one in the context of allowable insulation heating. Exceeding of allowable heating for a given insulation class brings about a so-called accelerated insulation ageing: the mechanical strength decreases, embrittlement, fractures, cracks and occur causing the insulation electric strength decrease and disruption.

Insulation resistance decrease. While IMs are in use their insulation is inevitably «ageing».
Principal causes of the ageing are heating of windings through operating and starting currents, short-circuit currents, overload currents and external source heat; through dynamic forces resulting from lead interaction and through switching surges. The insulation condition is influenced by ambient conditions such as air temperature and humidity, air pollution and dust content.
The degree of operational safety of electrical installation is determined by the insulation condition. An IM is allowed to be used only if its insulation resistance to frame is not less than
0,5 Mom. The insulation disruption probability increases by an order of magnitude if the
insulation resistance is twice less than the allowable one.

The insulation resistance drop increases the probability of such IM fault as the stator winding breakdown to the frame (ground fault) which is dangerous not only for the IM itself but also for the service staff. Short-circuit currents that 10-100 times exceed the nominal currents start to run through the supply line and the electric installation frame can be found under high voltage that is endangering life. It’s very significant to carry out continuous monitoring of the stator winding insulation resistance during the IM use because dielectric insulation properties measured before the IM cut-in can suddenly change under influence of electric voltage and temperature. For this purpose the ground leakage current is measured by the differential current transformer that responds to the occurring of the difference current that exceeds the user-specified setting.

Fault condition protection methods. In order to provide IM protection against
faulty conditions the different kinds of relay protection – thermal, current, temperature, filter and combined ones – have been come into use since the middle of the last century.

For the years of IM service experience it has been shown that most existing systems don’t provide the fail-safe operation of IMs. Thus, for instance, thermal relays are designed for long-term overload that is 25-30% of the nominal one. But more often they trip at one phase loss under the load that is 60% of the nominal one. Under a lesser load the relay doesn’t trip and an IM keeps operating from two phases that brings about the winding insulation overheating and the IM failure. The correct choice of the protection device is an important factor for the fail-safe operation of IMs.
Protection devices against IM faulty conditions come in several types:
A) thermal protection devices: thermal relays, tripping devices;
B) current-dependent protection devices: fuses, circuit-breakers;
C) heat-sensitive protection devices: thermistors, thermostats;
D) protection from mains faults: phase voltage & phase monitoring relays, mains monitors;
F) maximum current protection devices, electronic current relays;
G) combined protection devices.
In the second chapter of this article we are trying to give a detailed description of principles of operation, strengths and weaknesses of the existing protections as well as to discuss attempts to create universal protection devices for IMs.

Sorkind Michail, chief manager of NOVATEK-ELECTRO LTD (St. Petersburg)