Minutes of the B513 Planning Meeting, 30th May

Present: Roland Bachelard, Tony Cass, Anne Funken, Jukka Lindroos, Dick Minchin, Dave Underhill

1.             Air conditioning

  1. Existing pipework in B513 will support an extra chiller.
  2. The 750,000m3/hour airflow limit is not imposed by working conditions but by the need to keep at least one air conditioning unit free for redundancy. This airflow requires nine units in service with one on standby rather than today’s 8/2 configuration. Roland pointed out that units 5 and 6 are dedicated for desenfumage today and cannot both be on at the same time. Jukka will investigate the modifications that are required for this to be possible.
  3. It is difficult to measure the real cooling needs of offices because the airflow depends on the number of cooling units in circuit. The estimates provided at the last meeting would seem to be reasonable, however. However the figures presented at the last meeting do not include the UPS requirement—100kW.
  4. Having investigated the use of the space currently used for MG room air conditioning equipment, Jukka concludes that it is more feasible to use the MG room itself for the cooling units for the vault and use the MG cooling equipment area for equipment/storage. Any air conditioning units for this equipment/storage area would be separate but installed in the <current MG room with those for the vault. Removing the MG room cooling equipment must be costed.

2.             B513 Electricity Supply

2.1.                General Building Supply Issues

  1. Electrical supply in B513 is provided by two 2MVA 18kV/400V transformers that supply two coupled switch boards. Of these, one powers the machine room, the raised floor area in the vault and a 10kVA supply to B31 via the UPS. The second supplies power for air conditioning and the general building circuits. Both of these boards are controlled by the Emergency Power Off circuitry. There is a 3rd board, not controlled by the EPO, that powers emergency equipment but this is not available for IT use.
  2. To allow for maintenance work on these transformers, the total load must be less than 2MVA. This limits “useful” capacity to 1.2MW, or 1.8MW if all maintenance work is carried out during the winter when the power requirements of the air conditioning are at a minimum. Although the 1.8MW is close to the predicted, but perhaps arbitrary, 2MW requirements for the LHC era, it must be noted that the air conditioning load will increase with the equipment load and thus the 1.8MW available for equipment will decrease.
  3. In any case, although the current transformers will last beyond LHC startup they will certainly have to be replaced before the end of LHC. It might therefore be sensible to plan to rework the electrical distribution system by 2004. Planning horizons are such that no formal position has yet been taken on the date for the replacement of the B513 transformers; Anne will ensure that a formal note is made concerning their lifetime. It was noted that the evolution in safety standards means that any replacement transformers would have to be installed outside B513.
  4. Replacement options include two 3MVA transformers, three 2MVA transformers or the separation of air conditioning and general services from the machine room load. All of these options should be costed. This is a long-term action for Anne once there has been the formal recognition that the existing transformers must be replaced.
  5. The regulatory authorities do not consider that the Swiss and French power supplies are independent and CERN must provide a fully independent power supply for critical functions. A pending report is likely to suggest that the central diesel generation system should be modernised rather than moving to a system of localised backup generators. However, if no other changes are made then any updated central diesel generators would not be able to supply more than 1.6MW to B513.

2.2.                UPS Issues

  1. The guaranteed lifetime for the current UPS: equipment is 15 years. This is sufficient to cover the LHC startup but even the most optimistic estimates of the likely lifetime indicate replacement well before the end of LHC running. As for the transformers, therefore, it would seem sensible to replace the UPS by 2005.
  1. UPS systems generally consist of a “bypass unit” and “onduleurs”. If the bypass unit is sized appropriately and configured, generation modules can be added according to need—but it must be noted that the commercial lifetime of these modules is much lower than the guaranteed maintenance lifetime.
  2. A report following a recent examination of the current UPS batteries referred to an “explosion risk”. Roland Bachelard clarified that the risk is of an explosion of an individual battery. Although such an explosion (or rupture) would have an impact on the UPS capacity there would be no immediate impact on services and there is certainly no general risk to B513. Nevertheless, replacement of the batteries during 2001 is strongly advised.
  3. Any replacement batteries could be used to power modules of any newly installed UPS if voltage and current requirements are met and that all batteries for a given UPS module are homogeneous.

2.3.                Power Distribution

  1. The current switch board and Normabar installation limits power consumption in the machine room to 1MW. There are no Normabars installed in the barn and capacity in the vault is limited. Roland will provide cost estimates for a distribution board and associated Normabars. Installation in either the barn or the vault would take some 2-3 months.
  1. Additional redundancy could be provided with the provision of a switch board at either end of the Normabars. It is not clear that this would be cost effective, however.
  2. A Thomson factory in Grenoble is apparently configured for guaranteed 24hour functioning; a visit to this factory could be interesting.
  3. There was an opinion that it would not be possible to have “local” Emergency Stop systems for individual areas of the machine room—rather any local Emergency Stop would have to be for the machine room as a whole. If this is the case then there might be little to be gained from moving away from today’s general Emergency Power Off system. This should be discussed with TIS.

3.             AOB

  1. It was agreed that the next meeting would be held at 2pm on Tuesday 29th June; the main topic to be the electrical distribution arrangements.

4.             Actions

4.1.                Completed

Who

Action

Assigned

Completed

Jukka

Determine the air conditioning capacity required for B513 offices.

09/05/00

30/05/00

Jukka

Investigate the use of the MG room air conditioning machine room to house air conditioning units for the vault.

09/05/00

30/05/00

4.2.                To be done

Who

Action

Assigned

Jukka

Investigate changes required to run 9 air conditioning units simultaneously whilst providing redundancy for desenfumage.

 

Anne

Arrange quote for removal of the existing MG room cooling equipment.

30/05/00

Anne

Ensure formal recognition of the need to replace the B513 18KV/400V transformers before the end of LHC running.

30/05/00

Tony

Ensure UPS battery replacement is included in IT budget for 2001

30/05/00

Tony

Arrange discussions with TIS about the requirements for general or localised emergency power off systems.

30/05/00