RE: Solenoid Radiator valves

["Neil Ball" <neilball@xxxxxxx>]

I can see the problem now Kevin - as you say, most valves do not give
balan=
ced flow across their opening range, most have almost 100% flow as soon as
=
they are partially opened which gives very little controllability.

Just another thought, are you measuring room temperature with sensors? If
s=
o, have you considered controlling the boiler output based on heating
deman=
d (direct boiler compensation)? Most commercial systems use either outside
=
weather compensation control, heating demand compensation based on room
sen=
sors, or a combination. Here you reduce the effects of the poor valve
autho=
rity as instead of only allowing a very small amount of high temperature
wa=
ter into the radiator you allow much more at a reduced temperature thereby
=
increasing the overall level of control. Works very well with condensing
bo=
ilers as it maximises the time the boiler is in full condensing mode for
mu=
ch of the heating season. You'll need to add at least on pipe sensor and
us=
e the heating demand to calculate the required boiler flow temperature,
and=
control the boiler temperature through simple on/off switching. You might =
also have to add a hot water sensor to make this a zone so that can
influen=
ce this calculation as well and boost the boiler temperature when the
cylin=
der needs heating.

With regard to positioning raise/lower valves then it is not too hard -
you=
simply roughly position the valve based on room heating demand, and then p=
ulse the valve in small increments thereafter based on room response. As
lo=
ng as you measure the full stroke time of the valve it becomes a
time-based=
operation, and you can always use the full stroke time plus a margin whene=
ver the valve is moved fully open or closed to ensure it gets registered
to=
the extreme of it's travel. In reality you get pretty good positioning wit=
hout the expense of 0-10v control (bear in mind that most 0-10v actuators
s=
imply turn this signal back into a raise/lower signal internally to drive
t=
he motor anyway).

Let me know how you get on.

Neil.




-----Original Message-----
From: Kevin Hawkins [mailto:lists@xxxxxxx]=20
Sent: Monday, April 07, 2003 13:58
To: ukha_d@xxxxxxx
Subject: RE: [ukha_d] Solenoid Radiator valves


I have played quite a bit with the algorithms for the ON OFF control
=E2=80=
=93 which I have in about 4 rooms here. My problem comes because of the
hea=
t retention in the radiators. They are the very old fashioned cast iron
one=
s. Now as soon as the valve is opened they heat up very quickly (as the
wat=
er flowing past is already very hot) but they hold heat for a long time,
ca=
using the overshoot =E2=80=93 so I am faced with almost pulsing the valves
=
open for a very short time to only heat the radiator up a little. I even
pl=
ayed with a temp sensor attached to the radiator. Unfortunately this then
a=
lso gets complicated by the flow temp at the valve and the flow rate or
pre=
ssure =E2=80=93 both these being effected by other demands from TVR=E2=80=
=99s and zones. In one of the rooms the radiators are more modern meaning
t=
hey are far more efficient at radiating heat and hence cool much faster
=E2=
=80=93 this room I have managed to cater for the hysteresis far better
with=
an algorithm.=20

TBH I had intended all my valves to be proportional type 1 or raise/lower
t=
ype 2 below, but I had installed the traditional zone valves assuming I
cou=
ld just exchange the actuators at a later time for proportional ones
=E2=80=
=93 but you can=E2=80=99t get proportional or raise/lower exchange
actuator=
s for standard zone valves =E2=80=93 so I had to work around or fit new
val=
ves as well.

Neil =E2=80=93 when I said proportional I meant to include raise/=
lower in that category =E2=80=93 I think we are talking about three types
h=
ere aren=E2=80=99t we ??

1)       The type that is either fully open or fully closed (but takes a
fe=
w secs to move from one to the other)

2)       The type that you apply say 0-10V to and they are proportional ie
=
3V is 30% open

3)       The raise lower type =E2=80=93 you drive them with a succession
of=
pulses to move them from one position to the next ( or maybe even a timed =
voltage application) =E2=80=93 then they stay wherever they are left
=E2=80=
=93 if you drive them 2/3 open then that is where they stay. These take a
l=
ittle tracking and have to be calibrated against end positions
periodically=
=E2=80=93 I think these are the lowest cost of the 2 and 3 and what I will=
probably end up using.

There is a fourth type which is what most of my radiators use at the
moment=
=E2=80=93 it is a sort of =E2=80=98bodged=E2=80=99 TVR. Basically I have T=
VR=E2=80=99s with a capillary and remote sensor mounted in a small plastic
=
box. You set the required temperature on the standard mechanical rotary
TRV=
. However also mounted in the little sensor box is a small electric
element=
that applies heat to the sensor in proportion to the voltage applied to it=
. Hence you can set back that room by 2 degrees by applying 2V or by 5
degr=
ees by 5V etc. This works very well to regulate the room as the control is
=
effectively truly proportional =E2=80=93 you can also effectively turn a 
r=
oom ON or OFF this way. I am not sure these are manufactured anymore
=E2=80=
=93 they were made by Danfoss.

=20

I am really interested in this subject as I have yet to implement most of
m=
y control here =E2=80=93 so please do keep us updated on anything you find=
=E2=80=A6

=20

Kevin

=20

=20

-----Original Message-----
From: Dave McLaughlin [mailto:dave@xxxxxxx]=20
Sent: 07 April 2003 12:36
To: ukha_d@xxxxxxx
Subject: RE: [ukha_d] Solenoid Radiator valves

=20

Hi Kevin,

I am also looking to replace mine with solenoid or proportional valves.
Did=
you have any kind of hysterisis in your control loop? Wouldn't adjusting t=
his have taken out the oscillations?

Regards
Dave...
-----Original Message-----
From: Kevin Hawkins [mailto:lists@xxxxxxx]
Sent: 07 April 2003 01:17
To: ukha_d@xxxxxxx
Subject: RE: [ukha_d] Solenoid Radiator valves


You should also consider using a proportional valve rather than just ON/O=
FF.
I have some ON/OFF ones and I still get oscillation around the setpoint d=
ue
to retained heat (forgotten the tech word) within the radiators. (I was
basically using zone valves)

I have researched a few inline / TVR replacement options =E2=80=93 You ca=
n get
plain ON/OFF down to about =C2=A325 but proportional all seem to be circa=
=C2=A360
plus. There is a very low cost device (TVR replacement) that works on
thermal expansion =E2=80=93 basically you apply voltage and it heats and =
expands but
the response time is quite slow.

Kevin


[Non-text portions of this message have been removed]