A data centre's data hall should be clean when it is taken into use, and its cleanliness is largely decided during construction. Continuous measurement shows in every work phase whether dust is under control, and helps time the verification of the space's cleanliness before the equipment is installed.
Why dust is a particular risk for a data centre
In most data halls, the equipment is cooled wholly or partly with air that flows through it. During construction, dust settles in the ventilation system and on surfaces. When the cooling starts, the air carries it into the equipment, and it is hard to get out again. That is why a cleanroom class is recommended for data halls: the ASHRAE data centre guideline (TC 9.9) recommends keeping the data hall in class 8 of the ISO 14644-1 standard.
In Finland, the client can also require the construction cleanliness class P1. Its aim is that dust does not spread from the work area into clean spaces or the ventilation system.
Two ways to measure dust
Mass. On a construction site, dust is measured as mass, in micrograms per cubic metre of air (µg/m³). A Finnish dust measurement guideline from 2025 (Pölymittausohje 2025) requires a measurement interval of no more than five minutes. The main quantity is PM4, and its eight-hour average is used to assess the exposure of the workers. Loopshore took part in writing the guideline.
Count. Cleanroom classification counts particles. In class ISO 8, a cubic metre of air may contain at most 3,520,000 particles of 0.5 micrometres or larger and at most 832,000 particles of one micrometre or larger.
Loop One measures both from the same air: the mass concentrations PM1, PM2.5, PM4 and PM10, and particle counts per cubic metre in four size classes (≥0.5 µm, ≥1 µm, ≥2.5 µm and ≥4 µm).
Measuring in the different phases of construction
Dusty work. In demolition, drilling and grinding, dust is kept inside the compartment with negative pressure and air purifiers. Measuring points are needed both in the work area and outside the compartment, to see whether the dust stays where it belongs. Pressure difference measurement with Loop Delta shows whether the compartment stays under negative pressure. An exceeded limit can trigger an alarm by text message or email.
Interior finishing and building services. When some of the spaces are already finished, the measurement shows whether dust travels into the finished data halls. Work phases and observations are recorded as notes together with the measurements, so the cause of a rise can be found later too.
Final cleaning and ventilation start-up. The particle count should fall and settle. The continuous curve shows when the level has stayed low long enough and in which spaces it still rises.
Taking the space into use. If the cleanliness of the data hall is verified before the equipment is installed, this is done with a classification measurement using a particle counter according to the standard. A classification measurement tells the situation at one moment. Loop One shows whether clear rises appear between the measurements, for example on installation and delivery days, and whether the level recovers after them. The same monitoring continues during use, for example when filters are changed.
Accurate measurement and continuous monitoring together
A portable particle counter measures accurately, but one measurement tells one moment. Loop One is there all the time, but its accuracy is lower. Together they give the benefits of both:
- The baseline is measured with a particle counter next to the Loop One devices.
- A site-specific correction factor is calculated from the readings of the counter and the devices and set on the devices remotely. The factor corrects the readings from then on.
- The comparison is repeated when the work phase changes, because the factor applies to the dust and the concentration it was measured with.
- When a block is finished, the devices are moved to the next one.
This way the continuous curve matches the particle counter better in the dusty phases, and the official classification measurement can be timed for a moment when the curve shows no rises.
Practical arrangements
- The device uses its own encrypted 4G connection, so it is not connected to the network of the site or the data centre.
- The starting point for dimensioning in the dust measurement guideline is one device per 150 square metres of work or affected area, at a height of about 1–2 metres.
- On a stand, the device can be moved from place to place without fixings to the structures.
What continuous measurement does not replace
- The official ISO 14644-1 classification is always done with a particle counter and a procedure according to the standard. Loop One does not classify.
- In a clean data hall there are so few particles that Loop One's accuracy is not enough to tell whether the space is below the ISO 8 limit. The device shows the dusty work phases and clear rises.
- ISO 8 also has a limit for particles of 5 micrometres or larger (29,300 per cubic metre). Loop One does not count this size class separately: its largest size class is particles of 4 micrometres or larger.
- The handover cleanliness of class P1 is assessed from dust settled on surfaces and in the ventilation system. A device that measures the air does not see it.
Sources
- ISO 14644-1:2015 (maximum counts of class 8)
- ASHRAE TC 9.9: Gaseous and Particulate Contamination Guidelines for Data Centers
- Classification of Indoor Environment 2018 (RT 07-11299), construction cleanliness class P1
- Pölymittausohje 2025: Pölyn mittaaminen rakennushankkeissa (Measuring dust in construction projects), Tampereen Tilapalvelut et al., 31.10.2025
- Loop One user manual, loopshore.github.io/docs