Choose a partial discharge tester by defining the assets, access points and measurement questions first, then specifying the sensor inputs, frequency ranges, sampling and synchronisation, noise rejection, analysis and reporting the instrument must provide. No single instrument fits every asset, so the RFQ should qualify the system against each asset class you actually test.
Define the Assets and Access Points First
The asset list defines the sensor requirement. Switchgear surveys need TEV and ultrasonic inputs; cable circuits need HFCT inputs at terminations; GIS needs UHF sensor capability with the right coupling; transformers need a measurement chain that can be coupled conventionally or via accessible sensors. List the asset types, their access points and the measurement mode for each, because an instrument without the matching input is not an option regardless of its other features.
Also define the workflow: a handheld survey unit for routine rounds, a multi-channel instrument for a focused campaign, or a fixed monitor for continuous surveillance. The same vendor may offer different products for these roles, and the RFQ must state which role is being procured.
Separate the instrument classes clearly. A survey unit optimises speed, weight and battery life; a campaign instrument optimises channels, analysis depth and data capacity; a monitor optimises reliability, communication and alarm handling. Trying to satisfy all three roles with one product usually compromises at least one, so the RFQ should be explicit about the primary role and accept secondary capability only where it does not weaken the primary one.
Specify Sensor Inputs and Frequency Ranges
Specify the sensor inputs explicitly: TEV input, HFCT input, UHF input, ultrasonic (airborne and contact) input, and the frequency ranges each channel must cover. The frequency response should match the signals of the target assets: UHF inputs in the appropriate band for GIS, HFCT inputs covering the pulse spectrum of cable discharges, and ultrasonic inputs covering the audible and ultrasonic range for location.
Ask how many simultaneous channels the instrument supports and whether each channel can be configured independently. A multi-channel system allows comparison between phases or between sensors, which is one of the most effective ways to separate a real discharge from noise.
Check that the sensor connectors and interfaces are compatible with the sensors you already own or plan to use, and that the frequency ranges of the inputs cover the signals of the target assets. A change of sensor family between measurement campaigns breaks the baseline, so sensor compatibility across the programme is a procurement requirement, not an afterthought.
Sampling, Synchronisation and PRPD Requirements
Phase-resolved partial discharge (PRPD) analysis requires the acquisition to be synchronised with the mains voltage, so the instrument needs a phase reference input and a sampling rate sufficient to capture the discharge pulses in the frequency range of interest. Specify the sampling performance and the synchronisation source (mains, internal or external) that matches your site conditions.
The analysis software should produce PRPD patterns, amplitude and phase statistics, and trend views that allow the operator to compare patterns over time. Since PRPD pattern interpretation is central to separating discharge types from noise, the instrument’s analysis and export capability is a core requirement, not an accessory.
Also specify the acquisition depth and storage: how long a capture the instrument can record, how many captures it can store before download, and whether the raw data or only processed statistics are exported. For investigations that need a second opinion, the raw trace and pattern data must be transferable, because a reviewer cannot re-analyse data that was discarded by the instrument.
Noise Rejection and Field Verification Features
Field PD measurement fails without noise control. Specify automatic noise rejection, gating and filtering options, and the ability to measure the background at each sensor position. Multi-sensor comparison, where two or more sensors are used to confirm that a signal originates in the asset rather than the environment, is the most robust field verification, so the instrument should support it as a standard workflow.
Ask how the instrument handles interference at the test frequency and how the operator validates a suspicious reading on site. A demonstration in a live substation environment, or a documented field reference, is worth more than a list of advertised features.
The field verification workflow should be part of the training: how to take a background reading, how to place and re-place sensors to test attenuation, and how to use gating to exclude a known interference source. An instrument is only as effective as the operator’s ability to apply these controls, so the purchase should include the workflow training.
Portable Survey Versus Continuous Monitoring
Decide whether the requirement is a portable survey instrument or a continuous monitoring system. Portable units are used for periodic rounds and focused campaigns; they need long battery life, rugged construction and fast setup. Continuous monitors are installed on critical assets, record over weeks or months, and need reliable communication, alarm thresholds and remote access to the analysis platform.
Some programmes start with portable surveys to build the baseline and later install monitors on the assets that show activity. If that path is likely, check that the sensor families are compatible between the survey and monitoring products, so the baselines remain comparable.
For continuous monitors, specify the communication path, the alarm notification method and the data retention period, and confirm the power supply and environmental rating for the installation point. A monitor that cannot communicate reliably in the substation environment will silently lose the very trend data it was installed to collect.
Calibration, Data Export and Supplier Support
Non-conventional PD instruments measure relative quantities, but the measurement chain still needs defined verification: sensor condition, coupling and gain settings, and periodic calibration or verification of the instrument. Ask for the verification procedure and the traceability of the settings, and confirm the software exports the raw traces, PRPD data and trend records in a format your asset management system can consume.
Supplier support matters for interpretation as well as repair: ask about training, pattern-library access, remote analysis support and the availability of spare sensors and leads. The instrument is part of a measurement programme, and the supplier’s ability to support the programme is part of the specification.
Reporting should match your asset management workflow: consistent naming of assets, automated inclusion of sensor positions and settings, and export to the format your system consumes. Manual transcription between the instrument and the database is the most common source of errors in PD records, so automation in the reporting path is worth a specification line of its own.
A Buyer’s Technical RFQ Checklist
Close the procurement with a written technical RFQ that states each requirement as a verifiable line item. The checklist below is the minimum structure; expand it with your asset list and site conditions.
| RFQ item | Requirement | Verification |
|---|---|---|
| Sensor inputs | TEV, HFCT, UHF, ultrasonic as required | Channel list and connector types |
| Frequency ranges | Matching each asset’s signal band | Datasheet response curves |
| Channels | Number of simultaneous channels | Configuration sheet |
| Synchronisation | Phase reference for PRPD | Input specification |
| Noise rejection | Gating, filtering, background and multi-sensor workflow | Demonstration or field reference |
| Analysis | PRPD patterns, trends, export | Software demonstration |
| Calibration | Verification procedure and traceability | Documentation |
| Support | Training, spares, remote analysis | Service agreement |
Require the supplier to confirm each line item in writing and to state any limitation in the same document, so the comparison between quotes is made on an equal basis.
Score the responses against the asset list and workflow defined at the start, not against the list of features. A system that covers the assets you test, the access points you have and the noise environment you work in, with training and support to match, is the one that fits; the same system may be wrong for a different site.
Frequently Asked Questions
How many channels do we need?
Channels should match your measurement workflow: one channel is enough for a simple survey, while two or more enable phase comparison and multi-sensor confirmation, which are the most effective noise-control measures. Define the workflow first and let it set the channel count.
Why is PRPD required?
Phase-resolved analysis shows how discharge activity relates to the AC cycle, which is the primary evidence used to distinguish discharge patterns from noise and to classify the probable discharge mechanism. Without it, the operator cannot interpret the measurement defensibly.
How should the instrument be calibrated?
Non-conventional PD instruments measure relative quantities, so the verification covers the sensor, coupling and gain chain rather than an absolute charge. Follow the manufacturer’s verification procedure, keep the records, and confirm the settings are traceable and repeatable.
For the method-selection framework behind these requirements, see the partial discharge testing guide. When you need to match a tester to your asset mix and access points, review PD test equipment options and request a technical proposal against this RFQ.
Keep the RFQ and the evaluation evidence together after the purchase, because the same document becomes the acceptance test for the delivered system and the reference for every later upgrade decision.