To initiate a data logging project, begin with the end in mind. Consider the overall objective of the study and what questions are to be answered with the data collected. Determine the scope of data logging activities accordingly.
Collect as much data as necessary but no more than is needed. Key decisions include the length of monitoring period, the interval at which measurements are recorded, and the data parameters to be collected. Collect data at a suitable logging interval and over a sufficient time period that will include the full range of operation relevant to the specific project, inclusive of evening and weekend operation. What is appropriate will vary based on each individual situation. A minimum of two weeks of data allows for comparisons over the two weeks and the ability to identify operational anomalies.
It is often advantageous to simultaneously monitor systems or components that operate interdependently in order to obtain operating data that is coincident. As an example, when collecting temperatures for an economizer function analysis, in addition to monitoring all temperatures simultaneously for comparative purposes, it is beneficial to also track supply fan operating status
When deploying multiple loggers whose data points are interrelated, data analysis will be much easier if all of the loggers have synchronized clocks, and are programmed to start at the same time and to collect measurements at the same recording intervals. Integrating multiple data sets with measurements collected at varying intervals can be extremely challenging within basic spreadsheet programs. And this is also true of combining interval data with on/off status data which has no set interval period associated with it.
Universal Translator is a software tool developed by Pacific Gas and Electric Company available for free download at www.utonline.org. It is particularly powerful for analyzing large data sets, and for reconciling data from different or unsynchronized sources or data that has been recorded at varying intervals. It includes tools to filter data and calculate psychometric properties as well as an array of modules to analyze economizers, light and plug loads, equipment runtime, etc.
Keep safety considerations at the forefront when deploying loggers. Each individual situation will present its own inherent safety issues, with exposure to ladders, electrical circuits and rotating equipment typically among the top concerns. Consider loggers that operate wirelessly via Bluetooth Smart technology for deployment in hard-to-reach or limited-access areas.
Inform the building operators where loggers will be installed on site. Facility staff unaware of data collection efforts may suspect the devices serve some nefarious purpose and remove them.
Install loggers securely to prevent them from shifting or detaching during the study period, and label them with a project note or contact information. Often it is best for the loggers to be positioned such that they are as inconspicuous as possible. This helps prevent tampering with the loggers.
Careful note-taking to document logger deployment details is critical. Many loggers have been lost because of inadequate field notes.The notes should include the exact location, the measurement intent, and a specific identifying number (like serial number) for each installed device, as well as the date and time the loggers were installed and removed.
Power Monitoring Best Practices
Safety is paramount, and this is most particularly important when logging electrical power. A qualified, licensed electrician should perform initial installation and removal of these data recorders. During the data-collection period, logging equipment should be secured within electrical enclosures or electrical rooms, and installed power meters should never be accessible to building occupants.
Become familiar with logger and current transformer specifications and instructions for optimal placement to ensure the most accurate results.
It is recommended that all three phases at the main panel be monitored rather than making assumptions about balanced loads. It is our experience that the power draw on different phases of a three-phase load is rarely equal.
It is common to monitor power in 15-minute intervals, as this allows for easy correlation with data from utility meters. Shorter intervals may be advantageous in some situations when power is expected to vary greatly over short time periods.
Because most power meters will provide amperage, voltage, and power factor values in addition to power (kW), it is often preferable to record these values along with the total power to make sure each contributing variable is consistent and reasonable over the entire study period.
Before leaving the loggers for the duration of the monitoring period, ALWAYS verify proper installation of the logging equipment as well as correct configuration of the logger software. Look at the real-time data values being collected to ensure they are within reasonable ranges of expectations for the given load, and confirm that the logging equipment is set to monitor all of the points important for the study. During the logger installation period is the ideal time to determine that a current transducer is installed backwards or a voltage lead is not fully connected, or a data point is not included in the logger output, etc. It is NOT the opportune time to discover these issues a month later when the data collected is being analyzed, as it is often impossible to correct data from meters installed incorrectly.
When a Variable Frequency Drive (VFD) or electronic ballast is being monitored, it is critical to install the power-logging equipment on the line (utility)-side of this equipment. The frequency and voltage on the load-side of the VFD or electronic ballast are outside the allowable range for most power meters and will cause inaccurate results.
HVAC/R Monitoring Best Practices
Use various types of loggers and sensors within their specified operating parameters.As an example, many loggers are not weatherproof and must be protected from moisture or excessive temperature exposure. Others can be operated via Bluetooth Smart technology and are ideal for hard-to-reach or limited-access areas within a facility.
Ensure the sensor is registering temperature attributable solely to the parameter being monitored. Shield the sensor as necessary to avoid direct sun or other temperature sources that will potentially influence readings and result in erroneous data.
Maintain consistency in logger model or sensor type across the study to provide uniformity of accuracy ranges and response times.
Particularly if the project includes comparing temperatures or analyzing differential temperatures, it is critical to comparatively evaluate the sensors or probes prior to use in the study. One approach is to select only sensors and loggers that provide consistent results from a larger batch of equipment. Creating these “matched sets” of loggers and sensors may require that the devices record variation similar to what they will encounter during the actual study. In some cases this pre-study data can establish specific correction factors for each logger/sensor combination that will compensate for any accuracy discrepancies.
Temperature sensors are often designed for specific applications and conditions. For instance, a high-range thermocouple is likely a best choice for a boiler study, but it will not be as accurate for the bulk of other HVAC/R applications. And some sensors can be damaged if exposed to extreme hot or cold conditions.
Response time of the temperature sensor and reading stability are other considerations to be matched to the behavior of the system under study.
When measuring the surface temperature of pipes related to hydronic systems, it is often best to use conductivity paste when strapping the sensor to the pipe to ensure good readings. After attaching the sensor to the pipe, it is often wise to place foam insulation over the sensor to buffer it from the room temperature conditions.
Plug Load Monitoring Best Practices
Plug-load studies can be approached by either monitoring accumulated energy over the study’s duration or recording watts, amps, volts and power factor over time. Measuring accumulated energy is simple and easy but does not provide data on the operation of equipment at specific times. If this level of detail is required, it is best to use a plug load monitoring device with built-in time-stamped data logging capabilities.
Plug loads can be measured at the individual receptacle. This approach allows a study to focus on a single appliance, or if a plug strip is used, multiple appliances.
Plug-loads can also be measured at the electrical sub-panel where an individual circuit (breaker), combination of circuits or an entire panel can be monitored. Meters installed in electrical panels require the assistance of qualified electricians to insure that the meters are installed safely and accurately.
Lighting Monitoring Best Practices
For accurate and useful data, it is critical to strategically position the logger in the space to be monitored. When monitoring occupancy in conjunction with a lighting study, occupancy detection will be accomplished with either ultrasonic or infrared sensors, each with their strengths and weaknesses. Check logger specifications and instructions to ensure accuracy of occupancy detection, and optimize logger location so it will detect only occupancy within the space of interest and not in adjacent spaces. Orient the logger so the light sensor is facing the light to be monitored. Ensure that the logger will not be able to sense light from another source (another fixture on a separate circuit, or daylight sources) that will compromise the validity of the study.
To avoid logger tampering during the monitoring period, light loggers should be placed out of sight inside the light fixture being monitored.
Do make sure that light loggers are not so close to the light source that the device is damaged by the heat.
Many lighting loggers that provide analog data have poor accuracies and cannot be calibrated. These devices are appropriate for determining light status and gross variations in lighting operation, but should not be used for fine-tuning of lighting controls. It is best to use sensors with cosine and color corrections where high accuracy is required.
In addition to observing how much time the lights are on with no occupancy, it is also a good idea to look at the opposite condition: “How often is the space occupied when the electric lights are off?” In some spaces with good daylighting or conscientious occupants, forcing the lights to a particular schedule would actually INCREASE energy usage!
Environmental Monitoring Best Practices
While sensor placement is often one of the most important criteria for obtaining accurate, useful measurements in all data collection projects, it is especially so with environmental monitoring. Sensors can be subject to local concentrations of CO or CO2 for instance, and may not represent the ventilation needs of the entire space. Because of this, the use of multiple sensors is appropriate for most Indoor Air Quality studies. Measuring at many locations will insure that pollutants are being removed effectively from the all portions of a space.
For CO2 studies, it is best to have a sensor in each occupied zone. Because the sensors are accessible, it is possible to get high readings from building occupants breathing on the sensor. Alternatively, the CO2 sensor can be placed in the return air duct where it will be free from tampering.
Envelope Monitoring Best Practices
Envelope studies often require that the surface temperatures of building assemblies are monitored. It is important that data loggers are properly programmed for use with external sensors and that the temperature range of these sensors are appropriate for the expected conditions.
To improve the accuracy of surface temperature studies, it may be best to place a small piece of foam insulation over the sensor to buffer the reading from the room temperature conditions.
When measuring surface temperatures of glazing, the sensors must be located so they do not receive direct sun. If the sensor does receive solar gain, it will heat up from the solar gain and will not accurately report the glass surface temperature.
Surface humidity levels cannot be measured accurately, as it is impossible to place the probe on the surface and get a valid reading. For these studies the air temperature and humidity near the surface is adequate, and the humidity level or dew point at the surface can be calculated with a psychometric chart.
When measuring light transmittance at windows, the inside and outside illuminance must be measured. One approach to this study is to measure horizontal illuminance inside and outside the window and include adjustments in the transmittance calculations that factor in the dynamic nature of the solar angle of incidence to the glass as the sun changes position over each day. Another approach is to measure the inside and outside light levels normal to the sun at a window that is receiving direct sun. For the latter test, a one-time reading will suffice, but adjustments must all be made for solar angle of incidence.
Runtime Monitoring Best Practices
For useful and accurate motor runtime hours, appropriate placement of the data logger is particularly important. For motor loggers that track on/off status based on the detection of the equipment’s magnetic field, confirm that the logger’s placement enables it to register the distinction between on and off for the motor under study, and that it is not being erroneously impacted by the field produced by nearby equipment. Similarly, lighting status loggers need to be installed so that they register only the condition of the electric lights under study and not light from neighboring electric or daylight sources.
Taking detailed notes to document logger deployment details such as the time installed and also removed from the field ca prevent potential errors in final data analysis. In the case of the pre-implementation data of the fan in Figure 10, (page 15) the “off” condition exhibited by the motor runtime logger on 2/14 correctly reflects that no magnetic field is being deterred; accurate analysis of this data is based on knowing this “off” status corresponds to the removal of the logger from the equipment as opposed to an actual shutdown of the fan.
Current Monitoring Best Practices
Collect data across the full range of variable operation or to include each level of operation for multi-stage equipment.
For amperage studies, a fast sampling rate of one minute (or faster) is typically recommended in order to detect all variation of the load.
To increase the accuracy of amperage-to-power conversions, simultaneous power measurements can be collected while current is being continuously monitored. A regression can be used to develop a mathematical relationship between the varying amperage and its correlating power draw.