Optimal Energy Efficiency Evaluation in Induction Machines Driven by Adjustable Speed Drives under EN 50598-2 and IEC 61800-9-1 Standards

Fig. 2 illustrates affinity laws governing VT fan and pump applications. For a given fan or pump geometry, Eq. (1) defines how flow, static pressure (SP), and power vary with two IM shaft speeds from 

When the IM speed is reduced to 50% of the full speed, the mechanical power can decrease to approximately 12.5% of its original value, demonstrating the dramatic and inherent energy savings benefit of using an ASD, especially in VT applications.  

However, in an actual system, overall system efficiency drops significantly at reduced shaft speed and/or partial load torque due to a disproportionate increase in losses. The energy efficiency optimization algorithms implemented in this paper can help achieve additional energy savings in such a scenario. 

In Fig. 3, the operating conditions are 480V ASD input in linear Volts/Hz mode, at half speed with a step load change from no load to a partial load of 15NM at 4s. The dc link voltage, IM currents, speed, and load torque all settle in their respective and expected steady state equilibriums. There is enough IM output voltage maintaining the required load level.

The strategy for higher energy efficiency is to determine the voltage and frequency to be applied to the IM with minimum losses. There is a unique combination of Id and Iq current components that provides the minimum IM losses for each operating condition. 

The quadratic Volts/Hz operation is activated at 4s in Fig. 4, where the load requirement cannot be sustained. This is the situation in which the energy efficiency improvement cannot be achieved. The quadratic Volts/Hz algorithm is only effective in reduced load instances, limiting its usefulness in a dynamic operating environment.

Since the IM parameters are required in this FOC method, the results can be sensitive to parameter variation over operating conditions and time. The computation burden is the highest among all methods described in this paper. 

D. Scalar Energy Optimizing Volts/Hz Control Algorithm 

In linear Volts/Hz mode, the maximum IM efficiency is typically at full load with rated slip. In VT applications such as HVAC or pumps, light load operation at lower speed is common. Two phenomena occur: 

(1). The slip decreases further from its rated value as the load reduces; 

(2). Even though the full IM torque is not required, the full magnetizing current still exists. It produces excessive magnetic field and reactive current that contribute to IM core and winding losses, generating heat and wasting energy. 

In order to overcome the shortcomings of the linear Volts/Hz profile, an optimal approach can be implemented. Figs. 7 and 8 illustrate the energy optimizing Volts/Hz control concept. In Fig. 7, by keeping the same load torque Tr and speed Ȧr, while the IM voltage is reduced, the slip increases accordingly. 

In the case of quadratic Volts/Hz, it could accomplish a similar goal in terms of improving IM efficiency, except that there is a stability and optimization limitation as described in Section III.(B). Fig.8 demonstrates two degrees of freedom (DOF) in the energy optimizing Volts/Hz control: 

(a). Change IM output voltage (moving up and down); 

(b). Change slip (moving left and right). 

Very often, the slip can be increased to be near its rated value, and the voltage is adjusted to reduce the magnetic field and magnetizing current. The overall system can still meet the load requirements, and optimal energy efficiency is achieved while maintaining the IM system stability. 

The theoretical explanation for the energy saving capability in Figs. 7 and 8 is analyzed. The IM shaft output power is expressed in Eq. (6), while the IM torque and slip are defined in Eq. (7). In this energy optimizing Volts/Hz control mode, the magnetizing current is minimized. Thus, neglecting Id and im due to a large Rm in Fig. 1(b), the rotor current is derived from complex vectors as in Eq. (8). The output power is calculated in Eq. (9). Furthermore, neglecting insignificant terms in Eqs. (8) and (9), the load current is proportional to the stator voltage Vs. Concurrently, Pout is proportional to Vs squared, which impacts the power consumption most notably. 

Where Ȧrm is mechanical speed (rad/s), Ȧr is electrical speed (rad/s), s is IM slip. 

The flowchart of the energy optimizing Volts/Hz is shown in Fig. 9. The system can start with a linear Volts/Hz profile, then during operation, IM data such as voltages, currents, speed and power can be acquired. With the objective to maximize the system energy efficiency, while meeting load requirements and maintaining stability, the energy optimizing Volts/Hz algorithm based on Figs. 7 and 8 is activated. 

Other advantages of the scalar energy optimizing Volts/Hz control platform include: 

(1). Plug and play with ease of use: It does not require user intervention due to the use of automatic adjustment; 

(2). It is insensitive to IM parameter variation, as compared to the FOC based flux optimization method in Section III.(C); 

(3). It is computationally efficient. 

E. European EN 50598-2 Standard and International IEC 61800-9-1 Standard for Energy Efficiency 

EN 50598-2 [1] was proposed by CEMEP. It defines energy efficiency indicators for CDM such as an ASD and PDS which is formed by CDM and motor load. The standard includes a methodology to determine the CDM and PDS losses, assigning the IE and IES values, which apply to motor driven equipment from 0.12 to 1,000 kW (100 to 1,000 V). 

Efficiency can be determined by testing or in conjunction with elaborate models for calculating inverter and reference motor losses. International Electro-technical Commission (IEC) adopted this standard and published 61800-9-1 [2] as an international standard in 2017. The evaluation of energy efficiency algorithm optimization is quantified according to Fig. 10.  

• Depending on the CDM power rating, the load current and displacement power factor are specified as a hypothetical test motor; 
• There are eight operating points with a maximum frequency of 90% of rated value. It is chosen such that the CDM over-modulation region can be avoided; 
• The CDM IE class is determined by the relative losses at the point of 90% speed and 100% torque; 
• The relative losses at the other seven points are specified in the product documentation, which are independent of the IE class definition.

• There are also eight operating points with a maximum frequency of 100% of its rated value; 
• The PDS IE class is determined by its relative losses at the point of 100% speed and 100% torque; 
• IES2 is the best efficiency class as defined in Fig. 10, which is achieved if a PDS has <20% of the reference PDS losses; 
• While the CDM IE rating is stated on the nameplate, the PDS IES rating is provided in product installation and maintenance manuals. 

A reference motor is defined by mathematical equations and/or power losses, used as a basis for comparing with other motors. As such, it might not be an available product on the market, it might be a generally available product from all concerned manufacturers. It may be as simple as any motor which has the required voltage and power ratings, or the next preferred rating above that of the converter. An electronic load is a possibility to simplify testing. 

• In order to determine losses or efficiency classes of a complete system, the CDM user needs data from the motor manufacturer or reference values as given in 5.3.3 in [1]; 
• Efficiency classes for sinusoidal fed asynchronous induction machines are defined in EN 60034-30-1. The classification is conducted for rated output (Pn: 100 % torque, nn:100 % speed); 
• The losses for the reference motor are derived from 4 pole asynchronous motors, using the 50Hz (applies to 60Hz as well), IE2 efficiency values according to EN 60034-30-1; 
• The inverter introduces a multiplier of rHL due to additional harmonic losses: rHL=0.15 for motors with a rated output power up to 90kW, rHL=0.25 for motors with a rated output above 90kW. 

IV. EXPERIMENTAL RESULTS AND ANALYSIS 

A. Experimental Evaluation on Energy Saving Algorithms 

For the 20HP, 480V, 60Hz system, the ASD delivers power to the IM, which in turn drives a size 24½ blower fan [11]. Taking the linear Volts/Hz control as a benchmark, three energy saving algorithms: quadratic, flux optimization and energy optimizing V/Hz control, as described in Section III, are implemented for a half load testing. At each operating mode, five adjustable frequency points are selected in the range between 30Hz and 55Hz. At each combination of frequency and operating mode, power consumption, temperature, IM speed, static pressure and atmospheric pressure are captured. Using linear Volts/Hz mode as a benchmark, it is demonstrated in Fig. 13 that the scalar Volts/Hz based energy efficiency optimization has larger energy savings than the other two methods.

Based on the test results, the energy optimizing Volts/Hz control mode is selected as the candidate for further evaluation under EN 50598-2 and IEC 61800-9-1 standards. 

B. Experimental Setup in Accordance with EN 50598-2 Standard and IEC 61800-9-1 Standard 

The ASD input and output powers are measured to evaluate the CDM efficiency and its IE class, while the ASD input power and mechanical shaft power output are recorded to quantify the PDS energy efficiency and its IES class. Fig. 14 illustrates the overall measurement system architecture, with the boundary definitions of ASD as a CDM, as well as a combined ASD and IM to form a PDS. 

Due to the harmonic content of the pulse width modulation (PWM) output waveform produced by ASDs, it can be challenging to obtain accurate and reliable system losses and efficiency measurement. Leveraging capability of ASDs to carry out loss segregation tests is investigated in [9]. Although it has potential to make measurements when elaborate test equipment is not available, further agency validation and certification are required for its acceptance. In this paper, certified test equipment and procedures are used in all data acquisition. Fig. 15 shows the test equipment including the precision power analyzer of WT3000 from Yokogawa [12]. 

Table I is a subset of test load displacement factor between fundamental output current and fundamental output voltage at different points of operation in [1]. A reference IE2 IM of 4.6kW, 460V, 60Hz is chosen in order to meet the test requirement for compliance with the standards [1-2].

C. Experimental Results and Evaluation on IE Class of CDM 

The experimental setup consists of a 400V, 50Hz, 7.6A ASD, a 4.6kW, 380V, 50Hz IM and a load control system. Since the ASD rated current is 7.6A, the actual IM load sits at 3.8kW at full load. Applicable to the operating conditions in the experimental setup, Table II specifies the reference CDM losses that are used in the IE energy efficiency classification evaluation. Where PrM , Sr- equ , Ir-out , PL- RCDM(90,100) are the reference CDM output power, apparent power, current, losses at 90% speed and 100% torque, respectively. 

Fig. 16 shows the baseline linear Volts/Hz CDM energy efficiency contour map with variations of IM speed from 10% to 100%, and load torque from 25% to 100%. The highest efficiency (>97%) is at near full load and speed, and the lowest of 74.5% is at 10% speed, 25% load. 

D. Experimental Results and Evaluation on IES Class of PDS Both the ASD and IM losses are considered in the PDS. The reference IM meets the load power factor requirement in Table I. Additionally, Table IV defines the reference motor losses at the tested power level. The measured IM losses at the subjected operating point are 521W. Thus, the reference IM also satisfies the IE2 motor loss definition in Table IV, even after including the switching losses due to the ASD non-sinusoidal output waveforms. PL- RM(100,100) is the reference motor losses at 100% rated torque and speed. 

Fig. 17 depicts the energy efficiency surfaces comparison between the CDM and PDS. In the PDS, the highest efficiency (82.9%) is at near full load and speed, and the lowest of 31.5% is at 10% speed, 25% load. Fig. 18 illustrates the PDS energy efficiency test results with and without the energy optimizing Volts/Hz strategy enabled. The elevated surface demonstrates higher efficiency operations at light load and across a wide range of IM speed. The highest energy saving is 7.9% at 25% load and 50% speed. At near IM rated or very low speed such as 10%, as load increases, the energy saving could be less significant. 

As in the scenario for CDM, Table V specifies the reference PDS losses that are used in the IES energy efficiency classification evaluation. Where PrM and PL- RPDS(100,100) are the reference PDS output power and losses at 100% speed and 100% torque, respectively.