Sprinkler Systems – Field Testing
INTRODUCTION
The scope of this Technical Bulletin covers sprinkler systems designed to BS/ EN 12845 Standard and BS5306: Part 2; older 28th Edition FOC and American standards are not specifically included although the principles of testing remain the same.
An essential part of evaluating the effectiveness of a sprinkler system for insurance purposes is witnessing testing of the sprinkler control valves and water supplies to ensure the equipment operates satisfactorily and there is sufficient pressure and flow available. The operation of the system should also generate an alarm, which may be audible only, but preferably linked to the fire alarm system and signal to a remote alarm receiving centre.
The tests must be carried out by a competent person, this would ideally be by the sprinkler maintenance contractor but may be a member of the policyholder’s staff provided they are familiar with the operation of the sprinkler equipment and are willing to carry out the tests.
Although the tests are ‘witnessed’ and the operation of the equipment is by the competent person(s), it will be necessary to specify the tests that are to be carried out.
PRIOR TO TESTING
Prior to carrying out any testing it will be necessary to determine the pressure and flow requirements of the system.
Where the system water supplies are direct from the town’s main the requirements for the system are calculated and measured at the installation control valves.
Where the water supplies include fire pumps, there are additional pressure and flow requirements that also need to be recorded within the pump house.
The majority of sprinkler systems will signal their operation to the fire alarm and therefore also prior to testing, the fire alarm will require to be isolated and any monitoring service advised of the pending tests. Assurance should be sought that any diesel driven pumps have had their normal weekly checks for oil levels etc. The level of water in the sprinkler tank should also be checked prior to flow testing.
SPRINKLER SYSTEM DESIGN CRITERIA
Pre-calculated systems are simplest to identify, as they have pre-determined flow rates for each design density and type of system, wet, dry etc. High Hazard pre-calculated systems still require the pressure to be calculated, so design documentation for the system will be required to determine the pressure requirements prior to evaluating the system.
The flow requirements for pre-calculated systems are provided in tables at the back of this Bulletin.
Fully hydraulically calculated systems have no such pre-determined flows and are designed based on a minimum pressure and flow from the individual sprinkler heads assumed to be in operation during a fire. The test requirements for these systems are based on the results of hydraulic calculations carried out for the hydraulically most remote and favourable areas of operation. The fire pump is selected to have a curve that is at least 0.5 bar above the pressure requirement for the most remote area of operation when the water in the sprinkler storage tank is at its lowest level. The term Qmax is the point on the pump curve where the most favourable area of operation demand curve intersects the pump curve with the tank full. The pump is designed to be capable at operating at Qmax plus 10%.
For example, a fully hydraulically calculated system may have a requirement of 2,320 l/min at 4.6 bar for the most remote area of operation and 2,310 l/min at 4.0 bar for the most favourable and a Qmax of 3,010 l/min. Test flows should therefore include these flows or be able to be calculated from the curve plotted from the test results. The maximum flow should be Qmax + 10% (3,311 l/min). Refer to the typical System Curve at the end of this Bulletin for more information.
SPRINKLER BELL TESTS
These tests ensure the installation generates an alarm and in the case of installations which operate in the wet mode, exercises the alarm valve to ascertain if it is in good working order.
Most installations will have an individual alarm gong; however, it is acceptable for the system to have a common alarm gong or even an electrically operated alarm only.
To conduct a bell test, the pressure both above and below the sprinkler alarm valve should be recorded before the test, the bell test valve fully opened, and the time recorded between the valve fully opened and the alarm sounding.
Each water motor alarm gong should be sounded for no less than 30 seconds to ensure the installation provides a continuous alarm.
The time taken for the alarm to sound will range from a few seconds to a few minutes in the case of very large wet installations (such as found in shopping centres).
Here we are looking for trends. The alarm for a particular installation should take a similar time to sound each time it is tested. If the weekly test cards show there is a significant change, then further investigation will be required.
Dry type installations are installed such that the test line is taken from the underside (upstream) of the dry valve and therefore does not exercise the dry pipe valve and the alarm should sound almost instantly after the test valve is opened.
Care is required when testing alternate type installations as the majority are fitted with a 3 way cock which is required to be turned 90 degrees before the winter bell test valve is opened. This is to redirect the bell line to suit the winter test position. Turning the 3 way cock in the wrong direction will instantly trip the valve set into the wet mode.
TRIP TESTS
Installations which operate in the ‘dry’ mode require to be trip tested at least annually. Where possible trip tests should be witnessed as part of the sprinkler survey. There are two types of trip tests which are both carried out with the system initially set in the dry mode, one which is conducted at the valve set and involves closing the isolating valve on the riser above the control valve set then tripping the dry/alternate valve set. This test ensures the valve set will function, but does not ensure that water will reach the most remote sprinkler head or nozzle within a reasonable time.
With the installation in the dry mode, the second type of trip test is a remote trip test and involves opening the remote inspector’s test valve (located at the hydraulically most remote part of the installation) and recording the time taken from opening the valves until water is discharged from the test pipe. Dry and alternate installations are designed with a larger area of operation based on water reaching the test valve in not more than 60 seconds. A longer delay may result in sprinkler failure.
Deluge and pre-action installations should also be trip tested. If possible, full deluge tests should be carried out to ensure that the nozzles are not blocked and that the spray is discharging in the correct direction.
Pre-action valves are tested with the installation in the normal stand-by condition, then trip the pre-action valve to allow the installation to fill with water. When a state of equilibrium is reached (determined by pressure stabilisation in the installation), open the quick-acting test valve of the remote test facility and record the time taken to discharge water.
FLOW TESTS AT THE CONTROL VALVES
Some sprinkler installation control valves are provided with direct reading flow meters which allows the pressure to be recoded at a range of flows during the test, however most systems have the flow meter installed only within the pump house.
Flow tests at the control valves requires the pressure to be recorded at 2 test flows, this represents a fire in the hydraulically most remote location and the highest pressure requirement for the installation. The second test flow represents a fire in the hydraulically most favourable location and therefore the highest flow requirement.
Where no flow meter is installed at the control valves, a 50mm full bore flow test is carried out to provide an indication that the water supply pipework is not blocked, or an underground valve is only partially open. The system pressure should be recorded when the 50mm test valve is fully open and the pressure has stabilised.
Where control valves are grouped together it is only necessary to carry out flow tests on the most remote set.
FLOW TESTS AT THE PUMP(S)
Pressures are required to be recorded at a minimum of 2 or 3 test flows and ideally 4 or 5 test flows, so a graph of the results can be produced. In addition, the pressure with the pump running at ‘closed valve’ should also be recorded as this is used in the calculation to check pump initiation pressures (refer to Other Tests and Records below).
The fire pumps should be tested individually with the jockey pump isolated. The situation where both fire pumps are isolated at any time should be avoided i.e. the pump that has been tested should be made live before any other is isolated.
The pump house will be provided with a flow meter. There are various types, and most are direct reading and record the flow in litres per minute. The test pipe normally returns the water to the tank.
The test regulating stop valve is downstream of the flow meter and should be in the closed position prior to testing. In most cases there is also an isolating valve upstream of the flow meter which will need to be fully opened.
The test regulating stop valve should be gradually opened until water begins to flow and any stop cocks on the flow meter assembly should be opened.
The flow should gradually be increased until the first of the test points is reached and the pressure recorded on both the pump delivery gauge and on the test/start rig gauge. For each test point the engine revs should also be recorded for diesel driven pumps or the amps for electrically driven pumps. The test regulating valve should be further opened until the next test point is reached on the flow meter and the process repeated until all the test points have been recorded, then the test regulating valve can be closed, and the pump stopped.
Repeat this process with each fire pump. Diesel driven pumps require a manual start after the test to reset the controller.
The flow test point should include the system demand point and the pump design maximum flow, for example if it was a pre-calculated Ordinary Hazard 15m pump, the test flows would include 1100, 1350 and 2250 litres per minute. If it was a High Hazard fully hydraulically calculated system, they would include the flows for the most remote and favourable areas of operation plus Qmax and Qmax plus10%.
Once the testing is complete the valves should be returned to their normal position and all pumps including the jockey put back to their operational condition. Some test return pipes drain automatically through a small hole bored in the pipe outside the pump house others require to be manually drained.
Finally, the remote alarm panel should be checked to ensure the relevant signals were received.
OTHER TESTS AND RECORDS
In addition to the pump flow test, the pump ‘cut in’ pressures (and ‘cut out’ for the jockey pump) are recorded. The primary pump (and standby pump) should have a cut in pressure of not less than 80% of the closed valve pressure of the primary pump.
The jockey pump should be set to cut out at the closed valve pressure of the primary fire pump.
Most pump houses are remote from the installation control valves and it is not practical to connect the pump house sprinkler protection to one of the control valves sets. In this case the pump house sprinkler protection is taken from a connection to the sprinkler main within the pump house and controlled by a stop valve, flow switch and test valve. The flow switch should be tested by opening the test valve and recording the time taken for the ‘Fire in Pump House’ signal to appear on the remote alarm panel. Flow switches are fitted with an adjustable delay of up to 90 seconds.
The ‘hours run’ of a diesel engine pump should be recorded and checked against the test cards to ensure it is being run for 30 minutes per week. Electrically driven fire pumps should be run for 10 minutes per week; however, this does not need to be recorded.
Diesel driven fire pumps should also be tested for failure to start. This is carried out by holding the engine stop lever in the closed position while the pump is on demand. The automatic starting sequence should make 6 attempts to start the engine. Each cranking attempt should be of at least 10 seconds and no more than 15 seconds duration, with a pause of not less than 10 seconds or more than 15 seconds between each attempt. At the end of each cranking cycle the engine cranking speed should be not less than 120 rpm while power is applied. The starting device will reset automatically. The system should also switch over automatically to the other battery after each starting attempt. At the end of the 6 attempts the engine ‘failed to start warning’ should illuminate on the controller and the remote alarm panel. It is only recommended that these tests are carried out by the servicing company as they can put strain on the batteries, particularly if it is an older system.
PRE-CALCULATED SPRINKLER SYSTEMS REQUIREMENTS
Typical system curve for a fully hydraulically calculated system
