Access Security FAQ
As the term implies there are multiple security zones that a person must pass security checks at before passing through. These can include an external barrier and fencing to prevent access to car park, to outer door security guards and inner door bio-metric-type security readers.
When an access control system uses an electronic lock, the lock only opens for a time-specific period when a valid card is inserted into a card reader. Without the access card the electronic locking device will not open.
A major benefit of an electronic fob is that an intelligent entry control system can normally identify the presence of a security fob and allow access. The fobs may be zoned to only allow access to specific areas of a building. The downside is that whilst the electronic fob may be allocated to a person, it can be handed to others for access.
Cooling Systems FAQ
SEER stands for Seasonal Energy Efficiency Ratio and is a term used in air conditioning to compare units for their energy efficiency.
Duty of Standby is a configuration scenario where one cooling system is referred to as the ‘Duty’ system and the other as the ‘Standby’. The choice between which device acts as the Duty is known as the service. In normal operation, the Duty air conditioner is running and the Standby is does not run. This type of configuration can suit server room environments using either single or three phase mains power supplies and where there is a need for redundancy (N+X) in the cooling operation. Server rooms with three phase mains power supplies can install either a three phase air conditioning system or single phase systems with one cooling unit per phase.
R32 is a refrigerant with relatively low flammability but zero Ozone Depleting Potentual (ODP) and a lower Global Warming Potential (GWP) than traditional R22 or R410A refrigerants used in air conditioners.
R22 is a Hydrochloroflourocaron (HCFC) refrigerant used in air conditioners, process chillers and industrial plant cooling systems. R22 is an ozone depleting gas and EU legislation banned its stockholding and system topping up from 1st January 2016. As a result air conditioners using this refrigerant cannot be maintained and must be replaced.
Relative Humidity is the ration of partial pressure of water vapor in an air-water mixture to the saturated vapor pressure of water at a particular temperature. Relative Humidity is shown as a percentage and measured in: %rh. In a high temperature area the difference can be as high as 10% or greater leading to potential damage and corrosion of metals, electrical and electronic devices. A typical example would be Singapore and other tropical-type environments where the average relative humidity can be 80% or more. In a datacentre or server room environment it is important to maintain low humidity levels.
A refrigerant is a fluid used within a refrigeration or cooling system to transfer heat. The refrigerant is circulated and absorbs heat at low temperature and pressure and transfers this heat at high temperatures and pressure. Refrigerants are typically man-made fluorocarbon compounds but there are natural refrigerants including Ammonia, CO2, hydrocarbons, water and air. Within an air conditioning unit, the term ‘refrigerant charge’ is also used. This is the amount of refrigerant, measured in Kg within the cooling system.
A Precision Air Conditioning (PAC) system is one designed for cooling datacentre and server room environments rather than one designed for general building (homes, commercial offices and retail). Precision Air Conditioning systems typically offer superior design and reliability and have a high ratio of sensible-to-total cooling capacity (COP).
Latent Cooling Capacity (LCC) is the amount of energy added or removed from the air in order to increase or reduce the moister content (humidity) during the air conditioning process. Latent Cooling Capacity is measured in Kg/Kg of dry air.
The term ‘hydrocarbon refrigerant’ is also known as HC and refers to a family of chemicals that contain only carbon and hydrogen, that are suitable for use as a refrigerant in cooling systems. Typical examples would include:
- Propane (R290)
- Isobutane (R600a)
- The CARE range with zero Open Drip Proof (ODP) and a very low Global Warming Potential (GWP)
The equipment within a server cabinet can include servers, switches, routers and rack mount UPS systems. These devices tend to draw air ‘front-to-rear’ as a result of their design and fan operation. A hot-aisle/cold-aisle containment is an arrangement of server cabinets so that the front of the cabinets face each other (cold-aisle). Air conditioned (cool) air is drawn in from the front of the cabinets and exits the server cabinets into a hot-ailse. From the ‘hot-ailse’ the warm air is drawn back into the cooling system for cooling and re-circulation back through the cold-aisle.
HVAC is an industry acronym that stands for ‘Heating, Ventilation and Air Conditioning’. These three systems are often combined into one overall system within a datacentre building with specialists known as HVAC engineers. Air conditioning specialist suppliers may also be known as HVAC companies. The aim for an HVAC system is to provide a thermal comfort and acceptable air quality in an economically viable and energy efficient manner.
Critical cooling load refers to the load in a server room or datacentre environment of the usable cooling capacity. The critical cooling load does not include equipment in support areas outside the datacentre or server room floor.
Design and Build FAQ
The most commonly quoted temperature average for a server room is 20-25°C. Whilst the ASHRAE standard promotes a higher temperatures to reduce cooling costs the room temperature has to be comfortable for anyone working within it. Also if there are UPS systems with batteries in a server room or data hall, the batteries tend to age quickly in ambient environments above 30°C.
High Performance Computing is a way of clustering computing devices to achieve a far higher level of computational activity than can be achieved via single computer. HPC is a term commonly associated with Super Computers used to solve complex mathematical and engineering problems. The term is also used within the datacentre industry where servers are clustered.
Form factor is a term used to describe the physical size and shape of a piece of computer hardware. The term is common within the IT industry and is used to identify how a piece of computer hardware should be installed: rack mount or tower form factor. If the device has a rack mount Form Factor a height will usually also be quoted in ‘U’. In a 19inch rack mount format 1U= 1.75inches or 44.45mm.
A BEMS is a Building Energy Management System and is an extension of a traditional Building Management System (BMS). A BEMS is a computer-based and controlled system that monitors and controls energy usage within a building’s electrical and mechanical systems including lighting, critical power systems, critical cooling, HVAC (heating ventilation and air conditioning) fire systems, access and security. The purpose of a BEMS is to reduce operational costs and aid decision making when it comes to the optimising energy usage.
Efficiency is a measure of the conversion of inputs to outputs. The higher the conversion ration the higher the efficiency. In terms of energy efficiency within a server room or datacentre environment the input for electrical equipment is electricity. Efficiency measurements (in percentage % terms) measure how much of the input power (Watts) is used to generate the output. The difference between this figure and 100% is wasted energy (normally in the form of heat and noise). Heat loss in a datacentre adds to the overall requirements for air conditioning. Energy efficiency in a server room or datacentre also affects electricity usage (measured on kilo-Watt-hours kWh). The higher the kWh of electricity used the higher the electricity bill for the period.
PUE stands for Power Usage Effectiveness and is a measure of the energy efficiency of an IT environment such as a server room or datacentre. The PUE term and calculation was developed by the The Green Grid. The PUE formula is:
PUE = Total Facility Energy / IT Equipment Energy
The nearer the PUE is to Unity (1) the more energy efficient the datacentre. Some mega datacentre operators can achieve a PUE around 1.2. For a low PUE ratio high efficiency systems must be used including uninterruptible power supplies, air conditioning units, lighting and servers.
Most server rack cabinets support up to 300Kg of internal weight. This will depend on the type of cabinet and its height (measure in ‘U’) and whether the rack cabinet itself is designed for light or heavy-usage.
Floor loading is measured in KN/m² where KN stands for Kilo-Newton. A newton (N) is the International System of Units (SI) for measuring the derived unit of force. Server rooms and datacentres typically have raised floors under which cooling and wiring is routed. The floors are made up of floor tiles that sit on a network of pedestals which are in turn fixed to a concrete base.
The floor loading on a tile and the four pedestals on each corner is important when calculating how much weight a floor can support in terms of a server cabinet or local uninterruptible power supply or computer room air handling unit. Spreader plates can be used to reduce the weigh on a single point loading. Structural engineers provide the necessary measurements and calculations for floor loading design and assessments.
It is important when designing a server room installation with a raised access floor to ensure the point loading value is not exceeded. Standard gravity is 9.80665 N/kg and so to convert KN/m² to Metric Tonnes/m² the following formula can be used: kN/m² X 1Kg/9.80665N = Kg/m².
Electrical Works FAQ
All our electrical contractors and electricians have to be IEEE 17th edition approved and certified by one of the national electrical contracting associations: the ECA or NICEIC. On completion of any electrical works, our qualified electrical team provide the necessary electrical certificate and can update any onsite electrical works registers.
We try to prevent any disruptions to services during the electrical side of installation projects by agreeing a documented installations plan with Risk Assessments and Method Statements (RAMS). Our teams will start prepartory electrical cabling works during normal working hours and will ensure there is no downtime prior to an agreed cut-over period. At this set time, the server room or datacentre will be powered down by your IT team. Once we get the green light that it is safe for us to proceed, our electrical team will kill power to the relevant part of the room and complete the ‘live’ electrical side of the installation. Once we have completed our testes and restored power, our project manager will give your IT team the ‘green light’ to begin the power-up of your servers and IT equipment.
Your building will be connected to the local electricity supply via a building incomer which will normally provide the supplies to the main HV/LV (high voltage/low voltage) switchboard in your building. This switchboard will provide power feeds to other distribution and sub-distribution boards within your building. A distribution board may be positioned on a floor with a sub-distribution board providing a feed into a server room. Each board will normally have an isolator and ways into which rated circuit breakers can be provided for individual power feeds within the server room including PDUs for server rows, air conditioning, UPS systems and other related equipment. When we survey a building or room for electrical works we normally need access to all the sub-distribution board in the room where we install the kit and may need to follow this back to the distribution board it is fed from and even the main building income and HV/LV switchboard or LV switchboard.
Energy Efficiency FAQ
We provide a complete energy efficiency audit service. This can be booked through our projects team who can arrange a suitable time and date. During the visit, we will do a site walk through to gain an overall view of the project and to identify easy-fixes and the hardware systems that will require review. After the visit, we provide a detailed energy efficiency report with recommendations for improvement and where necessary training and system upgrades.
We do not charge for an energy efficiency audit, if we can easily book the review into the diaries of our project managers. If a client requests an outside-working-hours audit or one that will involve us traveling over 250miles, we may have to make a special booking and will quote a reasonable price for this.
We often find quick solutions to energy efficiency lie in critical cooling and air conditioning systems and the air flow path within the server room or datacentre. Typical issues are poorly sealed floor tiles or rack cabinets with missing or poorly fitted blanking plates. Other solutions to poor energy efficiency figures (leading to high PUEs) include not upgrading to the latest energy efficient solutions including file servers, cooling systems and uninterruptible power supplies.
When we look at a server room or datacentre in terms of energy efficiency, we are looking at loads on the demand-side of the electricity distribution infrastructure. Related terms include Demand Side Management (DSM) and Demand Side Response (DSR). Our objective is to find ways to improve the overall energy efficiency of the facility through behavioral changes, more efficient hardware use and where necessary hardware upgrade to more energy efficient infrastructure systems (uninterruptible power supplies and cooling systems). We can make recommendations for IT hardware upgrade but focus purely on the infrastructure and related systems.
Fire Suppression FAQ
The costs to recharge a discharged or partially discharged fire suppression system depend upon its design. We provide a recharge service and would need to know the type of system installed, the number of cylinders, their location and of course the fire suppression agent require. We would normally carry out a site survey before quoting as part of our risk assessment and method statement.
It is mandatory by law to test a fire suppression system fitted to any building. Within a critical environment like a server room or datacentre we would recommend one to two inspections per year for maintenance and testing. Routine inspection and preventative maintenance will ensure your system is fully operational and ready to help protect your server facility and the people working within it from the risks associated with a fire.
Your server room facility may include sprinkler system components. The sprinkler head will contain a bulb with a coloured liquid inside that acts as a plug to prevent water from escaping. Heat rises and as the temperature increases from the fire, the heat expands the liquid which expands and eventually forces the bulb to burst, releasing the water in the sprinkler system. There are different coloured liquids for different temperature triggers including: orange, red, yellow, green, blue, purple and black. Sprinkler systems cannot be activated by smoke or dust even though the bulbs themselves are designed to be very fragile. Even tampering can set them off.
Floor and Ceiling FAQ
Floor loadings are measured in KN/m². For computer room Raised Access Floors and/or non-ground floor basement areas it is important to check the floor loading to ensure the the floor structure can support the weight of the server rack or UPS system to be placed upon it. A three-phase 100kVA UPS with a 5m battery can have a total weight of 1200Kg or 1.2tonnes. Spreader plates can be used to reduce point loadings and if necessary floors can be reinforced with additional central pedestals and supports. The total loading should be verified by a structural engineer if there are any doubts.
Servers can be sensitive to static electricity and electrostatic discharges. The floor tile and system chosen should have dissipative and conductive characteristics to deal with electrostatic discharges (ESD). Special floor mats can be also be placed on top of existing floor tiles of this purpose.
Static electricity is generated via contact and separation. Airflow can cause static electricity generation when it is laden with particulates that are large, numerous and visible. In a clean room environment there is little chance for static electricity to be generated or build up. The opposite is true in dusty environments or in some cases sandy ones.
The answer depends on what the ceiling void is to be used for. This could include network cabling and power cable trunking and routing or air flow. At a minimum it is best to leave 0.5 to 1M of ceiling void for access and to assist with repair or ceiling tile replacement.
Hire and Rental FAQ
We can provide a range of load banks and temporary power cables from less than 50kW to several Mega Watts (MW). This is generally arranged for clients when commissioning a new UPS system or standby power generator or during period preventative maintenance works.
We do and the minimum hire period is generally from one week, up to several months. Temporary power solutions like this can be used to cover an on-site failure or short-term increase in IT load demands. UPS and generator hire is subject to quotation and site survey if the system is hardwired. Plug and play solutions can be hired directly against a quotation and/or purchase order.
We can provide temporary air conditioning solutions when you need a short-term cooling system on site to augment exiting cooling availability or have a failed system. This is common during peak temperature periods which can mean that we have limited availability from our hire stock if not pre-booked.
IT Room Cleaning FAQ
Yes as well as cleaning the server rooms we can provide a complete clean of any ceiling voids under raised-floor areas. Provide we can safe access, we can generally clean any part of a server room or datacentre area.
As part of our service we can provide Tak mats. These help to reduce the risk of infection from spreading through foot borne contamination. We provide TAK mats made from a resin impregnated cotton in a heavy duty frame from the entry/exit points within server room space.
We use regularly maintained and PAT tested cleaning equipment. The HEPA (High Efficiency Particulate Air) filters in our vacuum cleaners are rated at 99.97% efficient at 3 microns. In addition all the cleaning agents we use are COSSH compliant and anti-static dissipative. This means that they will not generate a potentially hazardous electrical charge and they can also help to ground potentially hazardous electrical charges.
IT Room Relocation FAQ
We provide a complete UPS removals service including UPS decommissioning, electrical disconnection and battery pack disassembly. It is very important to make sure the battery circuit is open/broken when moving a battery pack. If the batteries inside are connected, they can discharge if the battery pack terminals are touched or make contact with a metal surface. Once safe to move, our logistics team can then transport decommissioned UPS to its new location where we can reconnect and recommission, ready for use. Our specialist logistics team can provide protective floor coverings, pallets, stair-climbs and lifts and hoists. Every care is taken to prevent damage to the equipment being moved and its surroundings.
We provide complete insurance during removal and relocation. We work with a specialist logistics company to ensure your critical and sensitive IT systems are safely packed and transported and unpacked and installed at their new location.
Our specialist team can move your computer room and IT servers to an external location or internally within your building. As part of our services we can provide new data cabling and labeling as well as the installation and roll-out of new equipment.
Modular Containers FAQ
A microdatacentre (or micro-datacentre) is a self-contained managed environment (in a cubicle or container) housing server racks, cooling (air conditioning) and power (UPS system) connections. A microdatacentre is a smaller alternative to a fixed-building datacentre and allows operators to install datacentre-type processing power close to the point of need in an ‘Edge’ computing format. Once installed the operator or user simply installs and configures their IT servers into the racks,
The term ‘containerised building’ is used to refer to a shipping-type container that has been converted for use into a building. These are usually older 20foot or 40foot shipping containers whose inner and outer sides have been treated and re-coated (sprayed) for their use as server room or datacentre containers. Inside the container, the fit-out may include heating, lighting, air conditioning, server racks, UPS systems and electrical switchgear and connection points.
Containerised buildings are fully functional and fitted-out datacentres that can be deployed relatively easily and quickly to meet client demand. This can be for Disaster Recovery (DR) or as a temporary measure to provide additional computing power. Containerised buildings can also be used for generators and uninterruptible power supplies. The benefit is that the building itself can be housed close to the main facility or in remote areas and does not require a long build program. Some Mega-sized datacentre operators use a modular approach for datacentre expansion and use custom-made containers and fabricated metal buildings.
Modular datacentres can provide a quick way to boost local onsite computing power or provide server computing power in a remote region. They can be supplied for permanent installation or on a temporary hire basis. Whatever the time period, a full site survey and project plan is need for a modular datacentre roll-out plan.
Network Cabling FAQ
Most new installations are now Cat6 and due to the backward compatibility, Cat5 applications will also work on Cat6. The difference between the two generation is transmission performance and bandwidth from 100MHz for Cat5 to 250MHz for Cat6. If an installation uses Cat7 the bandwidth can be 600MHz and each pair of wires are individually shielded to protect from external noise and interference
The accepted standard is 90m maximum between points and to allow a further 10m for patch panel connections. A patch panel connections a data cabling port to equipment in the server cabinet. Typical equipment can include server ports, data switches, telecommunications switches, CCTV cameras and WiFi equipment.
This provides DC (Direct Current) power over a Cat5 or Cat6 network for DC power items such as desktop telephones and other network connected items. The maximum distance between PoE points should be less than 90m as power can be lost in transmission.
Project Control FAQ
We provide a free of charge project management meeting with subsequent follow-ups and a project installation plan to accompany our quotation. This can be as detailed as required and we can interface this with plans from your other site contractors or sub-contractors as required. As part of our project plans we provide a range of assessment (environmental, quality, health & safety) as required. A RAMS (risk assessment and method statement) is generated if a site specific one is required. Otherwise we work to our generic RAMS.
We provide most documentation free of charge if available direct from our manufacturers. General Arrange (GA) drawings and single line diagrams (SLDs) are provided on request and may be charged for dependent upon their complexity and the number required. We do charge for O&M manuals and 3D-modelling with pricing on request.
Racks and Cabinets FAQ
Typical server rack cabinets can weigh up 150-300Kg when fully loaded. The server cabinet itself may be of a light or heavy-use type.
We can supply either completely assembled server cabinets or semi-knock down kits to allow for ease of transportation and logistics on-site. Our logistics team will follow the project plan (previously agreed) to deliver your server racks and enclosures to site and position them. Where the cabinets are supplied as semi-knock down kits, our installations team can assemble these and move into position for you.
Server racks are designed for front air intake and rear cabinet exit. Cool or air conditioned air should be drawn into the front of the rack by fans within the servers themselves or within the server cabinets and this then exits as warm air out of the rear of the cabinet. In a data hall server racks are typically positioned in rows and the rows positioned so that the fronts face each to optimise cool air and hot air flows. This is a basic for of hot-aisle/cold-aisle containment. The server racks may themselves be within a separate containment known as a hot-aisle/cold-aisle container. For some server racks we can also offer rear cooling doors to help collect the hot exhaust air. The cooling doors fit onto the rear of the server cabinets and use chilled water pipes within the rear door. We can also supply additional cabinet fans for extra cooling capacity.
If your server cabinet is only partially populated and has perforated ventilation holes in the front and rear, you may not need additional cooling fan kits. Server manufacturers typically recommend a minimum perforation factor for the doors of the cabinets that would be used to house their servers and this is typically from 63% upwards; to allow adequate air flow from front to rear.
Recycling and disposal FAQ
We often find that existing UPS systems typically outlast their initial battery sets. As well as provide new UPS battery sets we provide a complete recycling service through specialist recycling centres. These partners will strip the batteries into their constituent components for recycling, reuse and disposal as required. Each delivery is recorded on a Waste Transfer Note and we can provide a disposal certificate as required. We can also collect and dispose of old UPS systems, generators, IT kit and air conditioners.
We provide a recycling service for old computers, servers, monitors, mice, keyboards and printers. We can also recycle computer room furniture and carpets. Most items in a server room or datacentre can be recycled. If you consider a server, there are metal cases, plastics, cable harness and other materials that can be reclaimed in an eco-friendly way for reuse.
We can provide a financial sum, credit note or discount following removal and disposal of your old systems and consumables. We charge for the removal part of the service which includes decommissioning, strip-down and transportation to one of our licensed recycling centres. Once there, the team will assess what can be reclaimed and this is generally based on a price per Kg of material. The rebate or discount can then be quantified.
Room Monitoring FAQ
IT equipment gives off heat, even when idle or lowly loaded. High-end computing relies on powerful servers and in a server rack the amount of power drawn can be from 5-15kW or even higher. Even if the overall efficiency is 90%, that can still lead to 1.5kW of heat to clear through the rack via the cooling system. If the temperature rises too high, IT servers can fail and there is the potential for a fire to break-out in an un-cooled server cabinet.
An increase in humidity is a sign that the air conditioning system is not optimised or failing. Monitoring humidity is therefore important. Rising humidity can also lead to corrosion and for electrical/electronic equipment can lead to potential short-circuits and fire risks.
We can supply and install complete Data Centre Infrastructure Management (DCIM) software packages and the proprietary monitoring software for the systems we supply and install. For some of these bespoke alterations or custom interfaces can be requested. We also supply a wide range of interface cards and plug-in modules for remote software integration including SNMP and MODBUS/JBUS, RS232 and USB2.0.
Service Contracts FAQ
We split our maintenance contracts into three types including 4 clock hours (the fastest response time), 8 working hours or 12 working hours. The 4 clock hours is a 24/7, 365-day a year maintenance contract. We have been known to offer up to 2 clock hours but this is only feasible if there are one or two local engineers, who are literally on the site’s door-step.
Can I order a one-off service inspection to assess a critical system for refurbishment or replacement?
We classify these types of visit as a ‘health check’ and generally insist on a health check when we are asked to maintain a UPS system or air conditioner which we have not supplied or maintained. A health check can be ordered inside normal working or outside working hours. We provide a complete report following the health check assessment, with a recommendation for refurbishment, maintenance or replacement.
We do offer preventative maintenance visits and inspections. Normally we offer one or two visits dependent upon the server room or datacentre equipment and the client needs. Modern UPS systems for example normally only require one inspection per year but standby power generators and critical cooling air conditioners may require two.
Site Surveys FAQ
We provide free site surveys for electrical works, building works, UPS systems, generators, cooling systems and other works within a server room or datacentre, if we can schedule the visit into a normal working hour day for our project managers and the visit does not involve more than 250miles in travel. If the visit is required out-of-hours or requires more extensive travel a charge may be proposed.
The initial site survey is generally completed by one of our project managers. They will request one of our certified and trained electricians or electrical contractors to attend site with them (or to visit separately) if there is a need for extensive on-site electrical works. The works themselves are only undertaken by a qualified electrician. The electrical side of a site survey, always commences from the room distribution board or building incomer (whichever is relevant) to assess size, capacity, last test date, age of installation and suitability for the future works.
Whilst our project management team can cover most aspects of a general site survey, sometimes we also ask a member of our logistics team to attend. This can be the case, when a crane lift (hire) is required onto a roof space or to put equipment in through a window space or a HIAB lift is required to place a standby power generator over a fence onto a concrete plinth.
UPS and Power Distribution FAQ
Power factor is the ratio of Real Power (W) to Apparent Power (VA) in an AC circuit. Power Factor corresponds to the phase-angle difference between the load drawn voltage and current waveforms.
Power factor is shown as a decimal number or percentage i.e. 0.65pF = 65% between 0-1pF and 0-100% respectively.
Power Factor formulae include:
Power Factor (pF) = Real Power (W) ÷ Apparent Power (VA) = CosØ
If we know the Power Factor and Real Power we can calculate:
Apparent Power (VA) = Real Power (W) ÷ Power Factor (pF)
If we know the Apparent Power and Power Factor we can calculate:
Real Power (W) = Apparent Power (VA) x Power Factor (pF)
In electrical usage terms, Watts is the unit of measure for the Real Power (also referred to as Active Power) dissipated or drawn by a connected load. Some Unity power factor rated uninterruptible power supplies use Real Power for their rating.
Real Power (W) = Supply Voltage (V) x Amps (A)
For three phase loads, the Real Power is calculated for each individual phase and the three results added together to give a final total three phase Real Power result.
Power protection devices like UPS systems are typically referred to in terms of VA, kVA or even MVA. This measurement refers to the Apparent Power drawn by a load and is calculated as follows:
Apparent Power (VA) = Supply Voltage (V) x Amps (A)
In the formula (V) is the Root Mean Square (RMS) of the supply voltage and the Amps is the current drawn by the load. Ideally it is important to measure the current drawn at start-up and after the start-up process has settled down to the a running load current. To measure the Apparent Power drawn by a three phase load the formula is applied per phase and then the individual VA results are added together to give a final total.
Most UPS systems are designed to recharge their batteries to 80% within a 24 hour period. This is achieved through the UPS having an internal charging system. External battery packs may also be fitted with their own individual AC powered battery charger. From 80% the batteries are trickled charged as the typical charging curve is non-linear. The simplest way as a rule of thumb to calculate the recharge time is to take the Ampere-hour (Ah) of the battery set used and divide by the charging current:
Recharge time = Ampere-hour / recharge current
Edge computing pushes data processing nearest to the device that needs the information to operate, rather than pushing the data back to a centralised datacentre. Edge computing is essentially a de-centralised approach. In terms of power protection to ensure resilience in an entire decentralised eco-system it is important to protect both the servers within the cloud datacentre and each data-processing point along the Edge computing network. Larger three phase UPS systems may be required for datacentres with far smaller uninterruptible power supplies required for small load servers, PCs and Internet of Things (IoT) devices.