Elevator and Escalator Consulting Engineers
Hydraulic Oil Temperatures
We at KJA recommend maintaining the hydraulic elevator oil at a temperature of 37°C (100°F) plus or minus 10 per cent.
We are concerned with two things: the change in oil viscosity with the change in temperature and the possible breakdown of the oil at high temperatures.
Hydraulic elevators use a positive displacement pump to move the elevator in the up direction. A positive displacement pump cannot be throttled without damaging it. To control the flow of oil and thus the speed of the elevator in the up direction, the pump discharge is bypassed back to the reservoir during acceleration, deceleration and levelling. In the down direction the pump is not used; the flow of oil from the cylinder to the reservoir is throttled to control the speed during acceleration, running, deceleration and levelling. Hydraulic valves control the bypass of oil during the up direction trip and the throttling of the return oil during the down direction trip.
The hydraulic valves operate indirectly. A pilot valve is actuated by an electric solenoid. This pilot valve allows oil to flow under pressure from the system (from the pump in the up direction and from the hoistway in the down direction) into a chamber that actuates the main valves that allow the oil to flow. Main valve actuation speed is controlled by restricting the pilot valve flow. This allows a smooth speed transition.
If the oil viscosity changes, the rate of flow of the oil in the pilot valves and the main valves changes. As the temperature rises the viscosity drops and the oil flows more readily. If there is a marked change in viscosity the speed control changes: the elevator may level faster, the elevator may overshoot the floor and the operation tends to become less reliable. The hydraulic valves are temperature compensated but only to a limited extent so a significant change in the oil temperature can lead to speed control problems.
We normally specify an oil with a viscosity index in excess of 100. The higher the viscosity index the less change in viscosity with temperature.
With a good quality oil (viscosity index greater than 100) and with the temperature compensated valves and with proper adjustment a reasonable compromise in the speed control settings can be achieved that will result in acceptable operation at the oil temperature ranges normally encountered.
The speed control should be adjusted to avoid prolonged slowdown and levelling in the up direction. During this operation much of the pump output is being bypassed back to the tank. This adds a lot of heat to the oil.
Although it is highly desirable to maintain the oil temperature within our specified limits, a variation of 20°C (from 37°C to 57°C) is not abnormal and can be tolerated with the hydraulic valves properly adjusted. To ensure reliable performance it is our recommendation that consideration be given to the installation of a heat exchanger if the oil temperature exceeds 60°C.
Oxidization of oil normally does not occur below 75°C but the rate of oxidization doubles for every 10°C above 75°C. So temperatures above 75°C should be avoided since permanent oil degradation may result. It should also be noted that the maximum temperature at the worm of the positive displacement pump in the reservoir will be higher than the temperature of the oil in the tank. An oil odour in the elevator cab often is an indication of oil breakdown.
In normal operation the hydraulic elevator will start in the morning with the oil at the ambient machine room temperature. During the first two to three hours of operation the oil will rise to its maximum temperature and maintain that temperature until late in the afternoon. The oil will then cool down over the next four to six hours. This is a very general picture of the temperature change; the actual change in temperature is, of course, subject to elevator use.
A black body radiates about 400 watts per square metre at a temperature of 65°C. A typical oil reservoir might have a surface area of 5 square metres. This would allow 2.1 kW heat dissipation. This compares to 5.6 kW generated by the elevator when it is running more or less continuously. However, the reservoir heat dissipation works well enough for most hydraulic elevators since hydaulic elevators are rarely expected to deal with intense traffic.
If, however, the elevator is used continuously the heat dissipated by the oil reservoir may be less than the heat generated by the elevator operation. This excess heat has to be removed from the oil and from the machine room. To do this we can increase the air conditioning in the machine room but, although it helps, this will rarely be completely effective. A better option is the provision of a heat exchanger.
It is not necessary that the heat exchanger be capable of dealing with all the heat generated. Normally with problem installations a decrease of 10% will do the trick so even a 3000 BTU unit could be considered. Nonetheless, it is obviously better to have a margin of error and a 7500 BTU unit is a reasonable minimum. Moving up to a larger unit (17,000 BTU) is not prohibitively expensive.
There are two types of heat exchangers: those that fit on the oil reservoir and those that are installed remote from the machine.
The first type of heat exchanger fits into the reservoir and has fans (normally muffin fans) that blow air through the exchanger radiators into the machine room. There are water cooled models but it should be noted that coolers using water from the municipal supply are banned in some jurisdictions.
With the units installed on the reservoir we are still left with the problem of getting the heat out of the machine room. This generally means installing air conditioning or fans.
A better option is a unit using a remote radiator (outside the machine room). This obviates the need for additional machine room cooling. These units require running oil lines from the reservoir to the radiator and back. As part of the heat exchanger an oil pump is provided to move the oil and a fan is provided to move air through the radiator. However, in implementing this solution care has to be taken to respect the applicable fire regulations.
The remote radiator is best mounted in the machine room wall. These walls are frequently of cinder block construction, often have a parking garage or similar space on the other side and can be readily cut and patched for the insertion of the heat exchanger radiator.
The remote radiator can, of course, be mounted in any convenient place since there are only two low pressure oil lines and some control wiring running to the unit from the elevator machine and controller. It may require a wire cage protection if it is installed in an area open to the public.
A power supply (115 VAC 20 amps) is required for the heat exchanger. Our preference is to have the supply provided from the elevator controller by the elevator company but this does require the supply of a 3 kVA 600/115 volt transformer. Alternatively, an electrical contractor can provide the power supply.
If it is difficult or impossible to install the oil cooler remotely or if the fan noise of the remote cooler is objectionable or if fire regulations prohibit such an installation, the cooler can be installed in the machine room and additional machine room air conditioning can be provided to get rid of the heat. One of the objections to extracting air from the machine room is that the air inevitably carries with it a slight amount of oil and this oil can be a problem.
Although the manufacturers of these units will sometimes speak of a “four hour” installation, in practice things are rarely that easy. A typical installation would require a crew for one day but some applications might require more than that.
When converting a buried hydraulic cylinder to a sleeved hydraulic cylinder some thought has to be given to heat dissipation Generally speaking, a hydraulic cylinder with a PVC liner will cause the oil to heat up more than a cylinder buried in the ground.
The hydraulic cylinder (essentially iron) has a thermal conductivity of 59 W/m K which means, as one would expect, that it would transmit heat fairly well. Assuming that the cylinder protective wrapping is bitumen based it has a thermal conductivity of 0.17 W/m K. The surrounding soil has a thermal conductivity of 0.15 W/m K. The PVC liner has a thermal conductivity of 0.19 W/m K. The only real difference between the cylinder with the PVC and the cylinder buried in soil is the air space between the cylinder and the PVC. The air has a much lower thermal conductivity: 0.024 W/m K. However, with the air free to circulate and with the temperature of the oil (and thus the cylinder) tending to stabilize after the first hours of operation, it would be expected that the temperature of the air would be equal to the temperature of the oil and the effectiveness of heat transfer would then depend upon the thermal conductivity of the soil in both cases. There would be, even under stable conditions, a temperature drop across the cylinder-sleeve gap with a consequent higher oil temperature than without the protective liner.
It should be noted that the speed control adjustments might be changed at the time of cylinder replacement or the elevator (this is less likely) might be used less or more pre and post cylinder replacement.
Of course, most of the heat generated is dissipated in the machine room rather than through the cylinder wall and increased or decreased cylinder heat transfer will not greatly affect the total dissipation.
It is therefore difficult to perceive any significant difference between the situation with the cylinder buried in the ground and the cylinder inside a PVC sleeve buried in the ground.
Another case that is worth considering is that in which the original installation had a cylinder installed in a steel caisson with only air between the cylinder and the caisson for part of the cylinder. In this case the cylinder could act as a “heater” transmitting heat by convection. The effectiveness of this would be variable, dependent upon the width of the space, the exposed length of the cylinder and the extent to which the caisson is open at the pit level. A “guesstimate” would put it in the order of 2.2 kW.
As can be seen, this could dissipate a substantial portion of the heat generated by the elevator so that a busy elevator with an “open” cylinder such as this might work acceptably but when converted to a cylinder with a PVC liner the oil could overheat under the same operating conditions.
A method of approximately calculating cylinder heat dissipation is as follows.
Assume a cylinder 10" (0.254 m) in diameter with a length of 36' (10.7 m). This would have a surface area of 8.5 square metres.
The equation used to express heat transfer by conduction (Fourier's Law) is expressed as q = k*A*dT/s
q = heat transferred per unit time (W)
A = heat transfer area (square metres)
k = thermal conductivity of the material (W/m K)
dT = Temperature difference across the material (°C)
s = material thickness (metres)
We will assume that the temperature gradient is a straight line from the cylinder to one metre away from the cylinder at which point the ambient temperature of the soil is reached. Assuming the soil temperature is 15°C and the cylinder temperature is 65°C (This is higher than normal and would represent an installation at the border line of potential problems) gives a temperature difference of 50°C over a one metre thickness of soil. The value of k for soil is 0.15. The area, from above, is 8.5 square metres.
Using these data for input we get a value for q of 64 W or 218 BTU per hour.
The calculation of heat loss from a partially exposed cylinder is as follows.
Transfer of heat by convection has a similar formula to the heat transfer by conduction.
Q = h*A*dT
Q = rate heat transfer
h = coefficient of convective heat transfer
A = surface area
dT = temperature difference
Assuming h = 50 W per square metre Kelvin(The coefficient for air varies between 10 and 100), the air around the cylinder is 30°C (This assumes that the air is to some extent heated by the cylinder), with 1/2 of the cylinder length (we assume the same cylinder as in Appendix B) exposed, the cylinder surface temperature is 65°C, and heat transfer efficiency of 30% (The efficiency is an estimate based on the extent to which the cylinder-caisson space is sealed at the pit floor. It can only be an extremely rough guess and could vary from 0% to 50% or more), we get a value for Q of 2.2 kW.
How much heat does a hydraulic elevator generate?
The hydraulic elevator is quite efficient in the up direction with an overall efficiency in the order of 90%. In the down direction, however, the efficiency is zero; all of the potential energy being converted to heat in the oil.
If we have a hydraulic elevator with a 30 hp (22 kW) motor (100 fpm, 2500 pounds) , we will have roughly 10% loss in the up direction and 90% loss in the down direction. If we run the elevator up three floors (say 33 feet) we have roughly 20 seconds up at 0.1*30 hp and 20 seconds down at 0.9*30 hp. This represents 12.4 Wh up and 112 Wh down. So each trip would generate 424 BTU.
As a comparison we can calculate the potential energy converted to heat in the down direction. Assume a total weight of 8,250 pounds.(Otis gives forces of 25,232 pounds at the cylinder head and 11,393 pounds at each buffer for a 2500 pound hydraulic) The potential energy converted to heat is 350 BTU.
None of these calculations is particularly accurate (they assume full load and a bunch of other things) but the significant and overriding factor is the number of trips. Between door close time, running time, loading time, unloading time and door open dwell time the average round trip would take about 80 seconds. So the maximum number of round trips per hour would be in the neighbourhood of 45. This would give a heat loss per hour of about 19,000 BTU or 5.6 kW. It is rare that a hydraulic elevator would run continuously for any length of time. More probable is something in the order of 30% to 40% of this as the average heat loss per hour over the working day.