Comparative Tests on

Differential Pressure
Control Valves
Final Report 56148/1 Edition 2
This report supersedes Final Report No: 56148/1
dated 10 October 2013

Carried out for

Danfoss A/S
Hydronic Balancing & Control

By Phil Stonard
22 October 2013
 Comparative Tests on
Differential Pressure
Control Valves
Carried out for:

Danfoss A/S
Hydronic Balancing & Control
Nordborgvej 81 A602
DK 6430
Nordborg
Denmark

Contract: Final Report 56148/1 Edition 2

This report supersedes Final Report No: 56148/1
dated 10 October 2013

Date: 22 October 2013

Issued by: BSRIA Limited

Old Bracknell Lane West,
Bracknell,
Berkshire RG12 7AH UK

Telephone: +44 (0)1344 465600

Fax: +44 (0)1344 465626

E: [email protected] W: www.bsria.co.uk

Compiled by: Approved by:

Name: Phil Stonard Name: Mike Smith
Title: Test Laboratory Manager Title: Director, Engineering
BSRIA Test

This report must not be reproduced except in full without the written approval of an executive director of BSRIA. It is only
intended to be used within the context described in the text.

© BSRIA Page 3 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES PREFACE

PREFACE
This report supersedes Report 56148/1 dated 10th October 2013

The following amendments have been made:

Page 1 and 3 Company title changed to Hydronic Balancing and Control

Page 5 Paragraphs 9 and 11 rephrased and expanded to achieve a fuller explanation.
Page 6 Company name changed to ‘Danfoss’ from the individual business unit name. Range of Valve
B corrected to 5-25kPa from 5-30kPa.

This change not does affect the overall results or conclusions of the report as originally issued.

© BSRIA Page 4 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES SUMMARY

SUMMARY
This report concerns tests conducted on six differential pressure control valves. The work was
commissioned by Danfoss and carried out during the period 6th April to 12th June 2013.

Danfoss supplied three of each of their own valves, one of each was selected by BSRIA for test.
Other valves were purchased from local suppliers by BSRIA as single samples.

The test criteria, methodology and format of result presentation was as requested by Danfoss.

The conclusions on the valve performance are summarised as the range deviation and control mid-
point deviation value stated in the table below, which is the defined here as the difference between the
mid-point of the control band over a flow range of 5%-90% Qmax10 actual achieved and the nominal
set control value (DPc) of 10kPa.

These conclusions are based on the assessment method agreed with the client to allow a comparison
using a common calculated datum flow percentage for all valves. This may not be the same as the
methods used by different manufacturers when producing their own data, which may involve
particular quoted Qnom at each setting.

It is important to be aware that a small mid-point deviation (vs nominal setting) does not necessarily
equate to a small range deviation. A valve that exhibits a large range deviation over the operating flow
range can achieve a small mid-point vs nominal setting deviation.

The ideal in this test scenario would be both a small range deviation and a small midpoint deviation
from the nominal required value of 10kPa which would show evidence of a flat control profile with
minimum hysteresis at a mid-value close to the setpoint.

The midpoint value achieved compared to the nominal ∆Pc value is a measure of setting accuracy
under these test conditions.

Table 1 Interpolated test data

Valve DPc kPa DPv kPa Midpoint kPa Range deviation kPa
ASV-P (A) 10 10-50-150 10.1 ±2.3
ASV-PV (B) 10 10-50-150 10.8 ±2.8
C 10 10-50-150 13.4 ±6.9
D 10 10-50-150 9.5 ±3.3
E 10 10-50-150 16.5 ±6.0
F 10 10-50-150 10.3 ±4.7

Note that Valves C and D did not show evidence of the valves being fully open (usually seen by
means of the fall of the chart plot line to zero) when tested using the method within this report. It was
initially considered that this could be an indication of the simulation valve being too small. However,
3 different simulation valves were tried over the course of testing and the results within this report all
use the final simulation valve selected which had a Kv of 6.4. This coefficient was 0.6 higher than the
largest DPC Kv tested, thus giving no obvious restriction that would limit the test valve flow rate. On
examining the test rig pipework and pump no conclusion could be determined as to why this occurred.

Not all the valves achieved 100% of nominal expected flow rate, based on the valve manufacturers
flow coefficient (Kv).

All valves did however achieve pressure control over 5%-90% of the Qmax10 value.

Using the above method of assessment, valve A achieved the smallest range deviation and midpoint
closest to the set value.

© BSRIA Page 5 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES CONTENTS

CONTENTS
1  INTRODUCTION........................................................................................................................... 7 
2  OBJECTIVE .................................................................................................................................. 7 
3  ITEMS SUBMITTED FOR TEST .................................................................................................. 7 
4  TEST METHOD ............................................................................................................................ 8 
4.1  Instrumentation ...................................................................................................................9 
5  RESULTS .................................................................................................................................... 10 
5.1  Valve A .............................................................................................................................11 
5.2  Valve B .............................................................................................................................12 
5.3  Valve C .............................................................................................................................13 
5.4  Valve D .............................................................................................................................14 
5.5  Valve E .............................................................................................................................15 
5.6  Valve F ..............................................................................................................................16 
6  CONCLUSIONS .......................................................................................................................... 17 

FIGURES
Figure 1  Water flow test rig schematic.................................................................................................. 8 
Figure 2  Danfoss valve installed on rig ................................................................................................. 9 

TABLES
Table 1  Interpolated test data .............................................................................................................. 5 
Table 2  Interpolated test data ............................................................................................................ 17 

© BSRIA Page 6 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES INTRODUCTION

1 INTRODUCTION
This report concerns tests conducted on six differential pressure control valves. The work was
commissioned by Danfoss and carried out during the period 6th April to 12th June 2013. Matjaz
Osojnik from Danfoss attended the BSRIA laboratory on 4th and 5th April to observe the test
procedure. The test criteria, methodology and format of result presentation were all as requested by
Danfoss.

2 OBJECTIVE
The objective of the work was to conduct comparative tests to determine the ability of the valves to
maintain a set differential pressure (DPc) across the circuit when subject to a changing circuit flow
rate at various valve differential pressures (DPv). The flow range over which this was assessed was
5% to 90% Qmax10, where Qmax10 is defined as the maximum actual flow achieved at 10 kPa
differential pressure (DPv).

3 ITEMS SUBMITTED FOR TEST

The items submitted consisted of 6 valves, all of size DN20. There were two products from Danfoss:

ASV-P (Valve A) fixed setting 10 kpa

ASV-PV (Valve B) range 5-25 kPa

The remainder were valves from other sources, identified by letter reference below:

Valve C- range 5-30 kPa

Valve D- range 5-35 kPa
Valve E- range 5-30 kPa
Valve F- range 10-60 kPa

Danfoss supplied three samples each of two valve models. One of each was selected by BSRIA for
test. Other valves were purchased from local suppliers by BSRIA as single samples.

Each valve was supplied with its own capillary tube and fittings.

Valves identifications were checked by reference to packaging and valve markings. They were not
dismantled to examine items such as springs.

Some of the valves had dual function of pressure control and positive shut off, by two different
operations e.g. setting key and handwheel.

Some valves had direct settings for pressure that could be aligned with a body mark. Others required a
number of turns to be applied or an interpretation of an integer scale stem marking in conjunction with
a manufacturer’s graph. In the case of the latter, a judgement was made on the integer setting required
that was nearest to the nominal 10 kPa required, as no specific set-point for 10 kPa was stated by the
manufacturer- for example ‘setting 2’.

© BSRIA Page 7 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES TEST METHOD

Valves were set using the instructions supplied with the valves. Reference was also made to the
manufacturer’s electronic documentation available during the test period from their respective
websites.

All valves had instructions included within the packaging. These varied from simple pictograms to
more comprehensive text documents.

Each valve came preset to a value which was not necessarily that required for the tests. It was not
assumed that this was correct as a datum and therefore each was operated over its full range by means
of the handwheel or key before setting to the required value.

4 TEST METHOD
Each valve was set up on the BSRIA water flow rig, as shown in Figure 1.

Figure 1 Water flow test rig schematic

DPCV - Differential Pressure Controller

P1, P2, P3 – Pressure measurement points

© BSRIA Page 8 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES TEST METHOD

Figure 2 Danfoss valve installed on rig

Simulation valve

Test valve

P1 P2 P3

The system was filled and vented (including the valve pressure capillary tube).

A Magflo meter was used to determine water flow rate.

Two pressure transducers were used to determine DPc (tappings1 and 2) and DPv (tappings 2 and 3).
These had a range of 0.5 bar (50 kPa) and 1.6 bar (160 kPa) respectively.

All instrumentation was connected to an automatic logging and control system. DPc, DPv and flowrate
were logged at a maximum interval of once per second throughout the testing.

The simulation valve, which used a 0-10V motorised actuator, was driven fully closed>fully
open>fully closed to confirm operation before each test.

The test performed was a continuous movement of the simulation valve slowly driving from fully
closed>fully open>fully closed over a total period of approximately 66 minutes. The pump was
automatically controlled via an inverter drive to achieve a constant DPv for the duration of this period.

Each valve was tested at 3 different DPv: 10 kPa, 50 kPa, and 150 kPa, giving a total test time of
approximately 200 minutes.

Each test was conducted at a DPc setting of 10 kPa.

Data was plotted of pressure difference (DPc) vs ‘normalised flow’ of Q / Qmax as a percentage, where
Qmax was calculated using the manufacturers valve flow coefficient (Kvs) with a differential pressure
of 10kPa. A plotted line for rising (opening) and falling (closing) of the simulation valve was
produced for each DPv. The performance assessment used Qmax10 as the denominator.

The raw data for each full test amounted to some 20,000 data lines for each full 200 minute test and
consequently has not been included in this report. It is retained in the electronic project file for
reference if required.
4.1 INSTRUMENTATION
Description Identifier Calibration due
Seimens Magflo meter 0-1 l/s 124 25/06/14
Differential transducer 0-0.5bar 1317 06/04/14
Differential transducer 0-1.6 bar 1318 07/04/14

© BSRIA Page 9 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES RESULTS

5 RESULTS
The results are displayed as charts showing the valve pressure vs flow rate (as a percentage of Qmax
achieved). This was plotted for rising and falling (opening and closing) of the simulation valve.

To obtain a relatively smooth curve from the very high number of test points a trend line using a
moving average was employed.

The results were assessed from the charts in respect of the following features over a flow range of 5%
Qmax10 to 90% Qmax10 achieved:

 Range deviation (DPc), minimum value to maximum value (kPa)

 Midpoint, based on the mid value of the test range deviation, and the nominal set value (kPa) of
the valve.

Using the trend line pressures were rounded to the nearest 0.1 kPa for the performance assessment.

The chart for Valve A includes an illustration of the method of assessment for a single DPv value.

Note that the chart scale show Qmax as a percentage of the nominal flow based on the valve Kv,
The bold rectangle on the chart was the range assessed using Qmax10 as the denominator. It shows that
not all the valves achieved 100% of nominal flow rate, based on the manufacturers quoted maximum
valve flow coefficient (Kv).

The performance assessment in the conclusions is based on the achieved Qmax10

© BSRIA Page 10 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES RESULTS

5.1 VALVE A

© BSRIA Page 11 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES RESULTS

5.2 VALVE B

© BSRIA Page 12 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES RESULTS

5.3 VALVE C

© BSRIA Page 13 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES RESULTS

5.4 VALVE D

© BSRIA Page 14 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES RESULTS

5.5 VALVE E

© BSRIA Page 15 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES RESULTS

5.6 VALVE F

© BSRIA Page 16 of 17 Final Report 56148/1 Edition 2

COMPARATIVE TESTS ON DIFFERENTIAL PRESSURE CONTROL VALVES CONCLUSIONS

6 CONCLUSIONS
The conclusions on the valve performance are summarised as the range deviation and control mid-
point deviation value stated in the table below, which is the defined here as the difference between the
mid-point of the control band over a flow range of 5%-90% Qmax10 achieved and the nominal set
control value (DPc) of 10kPa.

These conclusions are based on the assessment method agreed with the client to allow a comparison
using a common calculated datum flow percentage for all valves. This may not be the same as the
methods used by different manufacturers when producing their own data, which may involve
particular quoted Qnom at each setting.

It is important to be aware that a small mid-point deviation (vs nominal setting) does not necessarily
equate to a small range deviation. A valve that exhibits a large range deviation over the operating
flow range can achieve a small mid-point vs nominal setting deviation.

The ideal in this test scenario would be both a small range deviation and a small midpoint deviation
from the nominal required value of 10 kPa which would show evidence of a flat control profile with
minimum hysteresis at a mid-value close to the setpoint.

The midpoint value achieved compared to the nominal ∆Pc value is a measure of setting accuracy
under these test conditions.

Table 2 Interpolated test data

Valve DPc kPa DPv kPa Midpoint kPa Range deviation kPa
ASV-P (A) 10 10-50-150 10.1 ±2.3

ASV-PV (B) 10 10-50-150 10.8 ±2.8

C 10 10-50-150 13.4 ±6.9

D 10 10-50-150 9.5 ±3.3

E 10 10-50-150 16.5 ±6.0

F 10 10-50-150 10.3 ±4.7

Note that Valves C and D did not show evidence of the valves being fully open (usually seen by
means of the fall of the chart plot line to zero) when tested using the method within this report. It was
initially considered that this could be an indication of the simulation valve being too small. However,
3 different simulation valves were tried over the course of testing and the results within this report all
use the final simulation valve selected which had a Kv of 6.4. This coefficient was 0.6 higher than the
largest DPC Kv tested, thus giving no obvious restriction that would limit the test valve flow rate. On
examining the test rig pipework and pump no conclusion could be determined as to why this occurred.

Not all the valves achieved 100% of nominal expected flow rate, based on the valve manufacturers
flow coefficient (Kv).

All valves did however achieve pressure control over 5%-90% of the Qmax10 value.

Using the above method of assessment, valve A achieved the smallest range deviation and midpoint
closest to the set value.

© BSRIA Page 17 of 17 Final Report 56148/1 Edition 2