Liberating the energy consumption meter/monitor

Liberating the energy consumption meter/monitor

With the advent of deregulation in many countries and decentralisation in others, the metering industry has been buzzing at conferences and exhibitions worldwide with papers and talk about how electronics will enable data collection, handling, and management. Ultimately we all benefit with better energy management practices that reduce peak demand and control the number of new power generation plants.

Most residential meters still measure and accumulate watt-hours electro-mechanically. Some add an electronic board to attain prepayment, multiple tariff billing, data collection or net-working. Taking a meter out of service to refurbish it with an electronic board has proved its usefulness in the short term. For large scale deployment, an energy meter that incorporates an electronic energy measurement base greatly simplifies manufacturing and lowers cost as volumes increase.

A popular topic of discussion at the AMRA Symposium this year was the apparent ease of retiring our old friend, the electromechanical energy meter, and replacing it with a single integrated circuit and its current transducer. The same DC power supply required for the electronic board retrofit is suitable for operating an energy measurement device.

Fig: Actual Size of a single IC kWh measurement base unit on a CD-ROM case

A reference design was on display at the show to illustrate simplicity and verification of design (see figure 1). Some people seemed hesitant to accept this new technology. Unfortunately, few people recognised that periodic field calibration is unnecessary with electronics. An electronic meter that is accurate to 0.5% during installation will not drift with time.    Further issues of concern were compliance with the ANSI specification that requires a display on the meter, and long-term reliability. Surprisingly, these two concerns are closely related and could be remedied by considering new approaches, like reducing the number of components to increase reliability. For example, an electronic display adds unnecessary complexity to a meter.

An energy meter’s display belongs in the home

In North America, most residential energy meters are electromechanical; they use a series of dials to display kWh consumption. Most consumers either don’t know how to interpret these dials, or never read the meter to double-check a bill unless it is inconsistent with past bills. The residential consumer in North America has already found the external display on the energy meter useless. After all, hardly anyone complains that meter reads are now estimated. We don’t get the cards in the mail to fill out and send back to utilities anymore. Similarly, we do not expect a meter display of our telephone conversations – an itemised bill is sufficient. Furthermore, if we want caller identification then we pay for the additional component to display this information.

Fig 2 This is important because the cost of an electronic energy meter needs to be reduced. An electronic display that is robust across wide variations in environment and temperature can be costly when non-volatile memory, display driver electronics, and liquid crystal displays are considered. We should not try to make our electronic solutions into an electromechanical look-alike. Instead, an energy metering revolution can unfold when the displays outside our homes are recognised as unnecessary. Imagine an energy measurement pod placed away from your house – much like many of our natural gas meters, which are underground and out of sight.

A display of energy consumption either on a panel inside the home or available on a computer terminal allows consumers to exercise greater freedom in accessing data to make decisions about energy management and bill reconciliation. Electricity dispensing solutions such as CashPower™ (see figure 2) and PowerCom™ (see figure 3) allow energy consumption on credit and offer an LCD display inside the home as a user interface for paying bills. The small display increases consumer awareness of energy consumption and offers additional energy management tools such as load profiling. Web-based solutions such as the Transdata® Mark-V display energy consumption that is updated every 14 seconds to a computer terminal. It is fast enough for industrial use, so it should be more than adequate for residential consumers.

Fig 3

Feasibility demonstrated

Electronic energy measurement has already been demonstrated with highly functional three-phase meters. The three-phase meter market requires electronic solutions to meet higher accuracy to obtain power quality information and to operate industrial sites through net-work management efficiently. Three-phase meters use electronics to interface to current transducers and digitise the voltage and current waveforms. A high-speed, high performance micro-processor or DSP uses the digitised waveforms to calculate the necessary parameters.

The Transdata® MARK-V energy meter offers the most advanced features and functionality available in the market today (see figure 4). The MARK-V energy meter is deployed by over 160 utilities, energy service providers, and other companies worldwide.

The cost sensitive residential single phase market needs only a subset of the functionality of the electronic three-phase meter. For this high volume market, a converter that interfaces directly to the current and voltage transducers and instantly accumulates kilowatt-hours is a valuable building block for reaching innovative solutions. The residential market will soon attain accuracy, flexibility, and networking benefits of fully electronic solutions. Semiconductor manufacturers are helping to contribute to this goal by providing low cost, high performance building blocks that reduce power consumption.

Analogue Meters Have Drawbacks

The most popular energy measurement method is the analogue method of the electromechanical meter. An analogue solution is not synchronised to a clock or crystal but calculates the product of current and voltage continuously. Electronic solutions have been used in the past to duplicate the behaviour of a classic electromechanical meter. Examples of electronic analogue signal processing include Hall effect multiplication and discrete analogue multipliers.

The drawbacks of any analogue energy meter are numerous, as large energy consumers want better data to make energy management decisions and utilities want more data to improve the services they provide. Analogue meters lack stability; accuracy drifts over large variations in operating environment and long periods of time. Inexpensive electronic solutions have been attempted with analogue processing or time division multiplexing, but only digital processing meets all the requirements of a highly integrated, low cost energy meter. The choice in digital signal processing architecture will dictate time to market for the next generation meter, a meter’s unit cost, and its life expectancy.

Fig 4

Digitising Energy Measurement

The term digital implies a time sampled system. An ADC (Analogue to Digital Converter) samples the outputs of current and voltage transducers at a high frequency to translate a real world waveform to binary words that digital circuitry can understand and manipulate. The resolution and speed of the ADC will influence the quality of duplication of the continuous time signal varying at a fundamental frequency, and also affect the amount of microprocessor bandwidth required for calculations. The ADC requirements for energy metering are sufficient sampling rate, high resolution (i.e. 16 bits) low cost, and low power. First, a sampling rate of at least 2 to 4 kilo samples per second is required for the application. A basic rule of sampling theory states that the rate (frequency) of sampling must be at least twice the highest frequency content of the signal. This is called the Nyquist rate. The current ANSI and IEC specifications for energy metering require accurate measurement of frequency content up to the 20th harmonic (1kHz to 1.2kHz). Next, the resolution of the ADC needs to be high because of the relatively wide dynamic range and accuracy requirements (<0.5%) of the application. A lower resolution ADC is sometimes used, but variable gain stages are required to increase the effective resolution. This solution could prove to be less reliable and more costly than using a higher resolution ADC because of the extra components and development time required. The choice of ADC will impact on overall cost because of the price of the ADC itself, as well as the cost of external components supporting a general purpose ADC. For example, a general purpose processor with integrated converters requires additional components for amplifying and filtering the signals from the transducers (i.e. shunts or current transformers). Another approach is to increase the effective resolution of an ADC by using over-sampling techniques in conjunction with noise shaping and digital filtering. The digital circuitry can also include the circuitry needed to calculate the required metering information, e.g. Watts or VARS. This could reduce or even eliminate the MCU requirements in a design, and so significantly reduce cost and increase reliability.

Finally, the optimal ADC must not consume excessive power. One of the challenging aspects of electronic energy meter design is designing the power supply, which must be low cost and reliable for the life of the meter. Outside North America these supplies are often capacitor-based, as the cost of a transformer is often prohibitive. Capacitor-based designs have a low VA rating and are typically only capable of providing 10 to 15 mA at 5V. 

Information options

Once converted to a digital signal, the voltage and current waveforms can be multiplied, filtered and integrated by digital circuits to extract just about any information. Calculations such as active, reactive, and apparent power do not need to change. A fixed function DSP places values in registers to be read by a low bandwidth microprocessor. If intensive calculations such as those required for power quality monitoring are needed, a programmable DSP is appropriate.

Energy measurement is simplified from a spinning disk to a single integrated electronic component. A programmable DSP approach allows high performance three-phase meters with flexible methods of calculating common terms and opportunity to monitor various current and voltage quality parameters. The fixed function DSP ensures high accuracy in calculating common terms at a low cost while reducing the bandwidth requirements of a data management processor for billing and energy management. Environmental robustness and simplicity of design gained by separating the energy measurement from the display of kWh will lower cost, increase reliability, and improve customer service.

Trademarks and products listed herein are the property of their respective holders.

  1. CashPower is a registered trademark of Energy Measurement Limited (A Siemens and Spescom joint venture company),
  2. PowerCom is a registered trademark of Motorola, Inc., ® Reg. U.S. Pat. & TM. Off.,
  3. Transdata® Mark-V is a registered trademark of Transdata Inc.,