Hukseflux HFP01 User manual

HFP01 & HFP03
Heat Flux Plate
Heat Flux Sensor
USER MANUAL
HFP01/ HFP03 manual version 0612
Edited & Copyright by:
Hukseflux Thermal Sensors
http://www.hukseflux.com
e-mail: [email protected]

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 2/35
Contents
List of symbols 4
Introduction 5
1General Theory 6
1.1 General heat flux sensor theory 6
1.2 Detailed description of the measurement: resistance error,
contact resistance, deflection error and temperature
dependence 8
2Application in meteorology 11
3Application in building physics 14
4Specifications of HFP01 16
5Short user guide 19
6Putting HFP01 into operation 20
7Installation of HFP01 in meteorology 21
8Installation of HFP01 in building physics 22
9Maintenance of HFP01 24
10 Electrical connection of HFP01 26
11 Appendices 27
11.1 Appendix on cable extension for HFP01 27
11.2 Appendix on trouble shooting 28
11.3 Appendix on heat flux sensor calibration 29
11.4 Appendix on heat transfer in meteorology 30
11.5 Appendix on heat transfer in building physics 32
11.6 Appendix on HFP03 33
11.7 CE declaration of conformity 35

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 3/35
List of symbols
Heatflux ϕW m-2
Thermal conductivity of the
surrounding medium or object on which
the sensor is mounted λW/mK
Voltage output V V
HFP01 sensitivity Esen µV/Wm-2
Thermal conductivity dependence of Esen E
λmK/W
Time ts
Surfacearea Am
2
Electrical resistance ReΩ
Thermal resistance Rth Km2/W
Temperature TK
Temperature dependence TD %/K
Depth of burial d m
Subscripts
Property of the sensor sen
Propertyofair air
Property during calibration cal
Property of the object on which HFP01 is
mounted obj
Property at the soil surface surf

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 4/35
Introduction
HFP01 is the world’s most popular sensor for heat flux
measurement in the soil and through walls and building
envelopes. By using a ceramics-plastic composite body the total
thermal resistance is kept small.
HFP01 serves to measure the heat that flows through the object
in which it is incorporated or on which it is mounted. The actual
sensor in HFP01 is a thermopile. This thermopile measures the
differential temperature across the ceramics-plastic composite
body of HFP01. Working completely passive, HFP01 generates a
small output voltage proportional to the local heat flux.
Using HFP01 is easy. For readout one only needs an accurate
voltmeter that works in the millivolt range. To calculate the heat
flux, the voltage must be divided by the sensitivity; a constant
that is supplied with each individual instrument.
HFP01 can be used for in-situ measurement of building envelope
thermal resistance (R-value) and thermal transmittance (H-
value) according to ISO 9869, ASTM C1046 and ASTM 1155
standards.
Traceability of calibration is to the “guarded hot plate” of
National Physical Laboratory (NPL) of the UK, according to ISO
8302 and ASTM C177.
A typical measurement location is equipped with 2 sensors for
good spatial averaging. If necessary two sensors can be put in
series, creating a single output signal.
If measuring in soil, in case a more accurate measurement is
needed the model HFP01SC should be considered.
In case a more sensitive measurement is required, model HFP03
should be considered.
In case of special requirements, like high temperature limits,
smaller size or flexibility the PU series could offer a solution.
This manual can also be used for HFP03. Differences between
HFP03 and HFP01 are highlighted in a special appendix on
HFP03.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 5/35
Figure 1 Drawing of HFP01 sensor
Figure 2 HFP01 heat flux plate dimensions:
(1) sensor area, (2) guard of ceramics-plastic composite, (3)
cable, standard length is 5 m.
All dimensions are in mm.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 6/35
1General Theory
1.1 General heat flux sensor theory
As in most heat flux sensors, the actual sensor in HFP01 is a
thermopile. This thermopile measures the differential
temperature across the ceramics-plastic composite body of
HFP01. Working completely passive, it generates a small output
voltage that is proportional to the differential temperature that
powers the heat flux travelling through it. (heat flux is
proportional to the differential temperature divided by the local
thermal conductivity of the heat flux sensor).
Assuming that the heat flux is steady, that the thermal
conductivity of the body is constant and that the sensor has
negligible influence on the thermal flow pattern, the signal of
HFP01 is proportional to the local heat flux in Watt per square
meter.
Using HFP01 is easy. For readout one only needs an accurate
voltmeter that works in the millivolt range. To convert the
measured voltage Vsen to a heat flux ϕ, the voltage must be
divided by the sensitivity Esen, a constant that is supplied with
each individual sensor.
ϕ= Vsen / Esen 1.1.1
HFP01 is a weatherproof sensor. It complies with the CE
directives.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 7/35
Figure 1.1 General characteristics of a heat flux sensor like
HFP01.
When heat (6) is flowing through the sensor, the filling material
(3) will act as a thermal resistance. Consequently the heat flow ϕ
will go together with a temperature gradient across the sensor,
creating a hot side (5) and a cold side (4). The majority of heat
flux sensors is based on a thermopile; a number of
thermocouples (1,2) connected in series. A single thermocouple
will generate an output voltage that is proportional to the
temperature difference between the joints (copper-constantan
and constantan-copper). This temperature difference is, provided
that errors are avoided, proportional to the heat flux, depending
only on the thickness and the average thermal conductivity of
the sensor. Using more thermocouples in series will enhance the
output signal. In the picture the joints of a copper-constantan
thermopile are alternatively placed on the hot- and the cold side
of the sensor. The two different alloys are represented in
different colours 1 and 2. The thermopile is embedded in a filling
material, usually a plastic, in case of HFP01 a special Ceramics-
plastic composite. Each individual sensor will have its own
sensitivity, Esen, usually expressed in Volts output, Vsen, per Watt
per square meter heat flux ϕ . The flux is calculated:
ϕ = Vsen/ Esen.
The sensitivity is determined at the manufacturer, and is found
on the calibration certificate that is supplied with each sensor.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 8/35
1.2 Detailed description of the measurement: resistance
error, contact resistance, deflection error and
temperature dependence
As a first approximation, the heat flux is expressed as:
ϕ= Vsen / Esen 1.2.1
This paragraph offers a more detailed description of the heat flux
measurement. It should be noted that the following theory for
correcting deflection errors and temperature dependence is not
often applied. Usually one will work with formula 1.2.1, possibly
corrected with 1.2.2.
When mounting the sensor in or on an object with limited
thermal resistance, the sensor thermal resistance itself might be
significantly influencing the undisturbed heat flux. One part of
the resulting error is called the resistance error, reflecting a
change of the total thermal resistance of the object.
Figure 1.2.1 The resistance error: a heat flux sensor (2)
increases or decreases the total thermal resistance of the object
on which it is mounted (1) or in which it is incorporated. This can
lead either to a larger of smaller (increase of or decrease of
the- ) heat flux (3).

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 9/35
Figure 1.2.2 The resistance error: a heat flux sensor (2)
increases or decreases the total thermal resistance of the object
on which it is mounted or in which it is incorporated. An
otherwise uniform flux (1) is locally disturbed (3). In this case
the measured heat flux is smaller than the actual undisturbed
flux,( 1).
A first order correction of the measurement is:
ϕ= (Rthobj+Rthsen ) V sen / E sen Rthobj 1.2.2
This correction is often applied with thin or well-isolated walls.
Note: this correction can only be determined for objects with
limited (finite) dimensions. For this reason this correction is not
applicable in soils.
In addition to the resistance error, the fact that the thermal
conductivity of the surrounding medium differs from the sensor
thermal conductivity causes the heat flux to deflect. The
resulting error is called the deflection error. The deflection error
is determined in media of different thermal conductivity by
experiments or using theoretical approximations. The result of
these experiments is laid down as the so-called thermal
conductivity dependence Eλ. The order of magnitude of Eλis
constant for one sensor type. For HFP01, Eλis given in the list of
specifications.
Esen = E sen, cal (1+Eλ(λcal - λmed)) 1.2.3
Note: this correction can only be applied when there is a
substantial amount of (at least 40 mm) medium on both sides of
the sensor. In soils λmed usually is not known. The value of λcal
typically is zero.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 10/35
Figure 1.2.3 The deflection error. The heat flux (1) is deflected in
particular at the edges of the sensor. As a result the
measurement will contain an error; the so-called deflection error.
The magnitude of this error depends on the medium thermal
conductivity, sensor thermal properties as well as sensor design.
In addition, the sensitivity of heat flux sensors is temperature
dependent. The temperatre dependence TD reflects the fact that
the sensitivity changes with temperature:
Esen = E sen, cal (1+TD (Tcal - Tsen )) 1.2.4
Combining 1.2.3 and 1.2.4:
Esen =
E sen, cal {(1+Eλ(λcal - λmed))+ (1+TD (Tsen - Tcal ))} 1.2.5
This correction is rarely applied because TD is typically small.
Apart from the sensor's own thermal resistance, also contact
resistances between sensor and surrounding material are
demanding special attention. Essentially any air gaps add to the
sensor thermal resistance, at the same time increasing the
deflection error in an unpredictable way. In all cases the contact
between sensor and surrounding material should be as well and
as stable as possible, so that it is not influencing the
measurement. It should be noted that the conductivity of air is
approximately 0.02 W/m.K, ten times smaller than that of the
heat flux sensor. It follows that air gaps form major contact
resistances, and that avoiding the occurrence of significant air
gaps should be a priority whenever heat flux sensors are
installed.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 11/35
2Application in meteorology
In meteorological applications the primary purpose is to measure
the part of the energy balance that goes into the soil. This soil
heat flux in itself is in most cases of limited interest. However,
knowing this quantity, it is possible to “close the balance". In
other words, apply the law of conservation of energy to check
the quality of the other (convective and evaporative) flux
measurements. For more information on meteorological
measurement of heat flux, see the appendix.
Users should be aware of the fact that the soil heat flux
measurement with HFP01 in most cases is not resulting in a high
accuracy result. The main causes are:
1 the fact that measurement at one location in the soil will have
only limited validity for a larger area; variability of soil surface
can be very lage.
2 the fact that variations in soil thermal properties over time
result in significant measurement errors.
If measuring in soil, in case a more accurate measurement is
needed the model HFP01SC should be considered.
Figure 2.1 Typical meteorological energy balance measurement
system with HFP01 installed under the soil.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 12/35
In a perfect environment, the initial calibration accuracy of heat
flux sensors is estimated to be +3 /- 3%.
In field experiments it is difficult to find one location that can be
considered to be representative of the whole region. Also
temporal effects of shading on the soil surface can give a false
impression of the heat flux. For this reason typically two sensors
are used for each station, usually at a distance of 5 meters.
Apart from the question of representativeness of the
measurement location, the main problem with heat flux
measurements in meteorology is that the sensitivity of a heat
flux sensor is dependent on the thermal conductivity of the
surrounding medium. This deflection error is described in chapter
1. In soil heat flux measurement the accuracy of soil heat flux
measurements very much suffers from the fact that the
surrounding medium is both unknown to the manufacturer and
changing over time. A typical HFP01 has a thermal conductivity
of 0.8 W/mK, while soils can vary between extremes of 0.2 and
4 W/mK. Sand in relatively dry condition can have a thermal
conductivity of 0.3 W/mK (perfectly dry 0.2) while the same
sand when saturated with water reaches 2.5 W/mK. A typical
HFP01 performing a correct measurement in dry sand will make
a – 16% error in wet sand. As in wet sand the heat tends to
travel around the badly conducting sensor, the flux will be
underestimated by 16%.
This example serves to illustrate that in soils where conditions
vary the so-called thermal conductivity dependence leads to
large deflection errors.
The third important error is temperature dependence.
Over the entire temperature range from -30 to + 70 degrees C,
the temperature error is +/- 5%. Taking the worst case soil,
pure sand, for the conventional heat flux measurement in
meteorology the overall worst case accuracy is estimated to be
+8 /- 24%. This is rounded off to +10 / -25 %.
In most situations the soil will not be pure sand, and in an
average climate the difference between the yearly extremes
might be -10 + 40 degrees C and a thermal conductivity range
from 0.2 to 1 W/mK. The temperature error then is +2 / -3%,
the thermal conductivity accounts for +0 / - 7%, the calibration
+3 / -3 %. The result of +5 / -13 % is rounded off to +5 / -15%

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 13/35
Heat flux sensors in meteorological applications are typically
buried at a depth of about 5 cm below the soil surface.
Burial at a depth of less than 5 cm is generally not
recommended. In most cases a 5 cm soil layer on top of the
sensor offers just sufficient mechanical consistency to guarantee
long-term stable installation conditions.
Burial at a depth of more than 8 cm is generally not
recommended, because time delay and amplitude become less
easily traceable to surface fluxes at larger installation depths.
See the appendix for more details.
Summary:
In case of use of HFP01 in meteorological applications, the use of
2 sensors per station is recommended. This creates redundancy
and a better possiblity for judging the quality of the
measurement accuracy. Typically one will work with two
separately measured sensor outputs; the average value is the
measurement result.
In normal soils (clays, silts) the overall expected measurement
accuracy for 12 hr totals is +5 / -15%.
In case of pure sands the overall measurement accuracy for 12
hr totals is +10 / -25 %.
The accuracy mainly is a function of the thermal conductivity of
the surrounding medium. In case of soils the moisture content
plays a dominant role.
The wider accuracy range in sand is due to the fact that the
thermal conductivity of sand varies with moisture content from
roughly 0.2 (perfectly dry) to 2.5 (saturated).
With other soils and walls (see chapter on building physics) the
variation of thermal conductivity is much less; roughly from 0.1
to 1 W/mK.
If measuring in soil, in case a more accurate measurement is
needed the model HFP01SC should be considered.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 14/35
3Application in building physics
HFP01 can be used for in-situ measurement of building envelope
thermal resistance (R-value) and thermal transmittance (H-
value) according to ISO 9869, ASTM C1046 and ASTM 1155
standards.
When studying the energy balance of buildings, heat is
exchanged by various mechanisms. The total result is a certain
heat flux. The dominant mechanisms are usually radiative
transfer by solar radiation and convective transport by flowing
air.
In most applications in building physics the sensor HFP01 is
simply mounted on or in the object of interest (see figure 3.1).
At the sensor surface, the convective heat of the air and the
radiation by the sun are transformed into conductive heat.
In case of incorporation into the wall, the conductive flux
through the wall is directly measured.
If direct beam solar radiation is present, the solar radiation is
usually dominant. The maximum expected solar radiation level is
about 1500 W/m2. In case of convective transport of heat by the
air, the convective transport is roughly proportional to the
difference in temperature between wall and air, and strongly
depends on the local wind speed. See the appendix on heat
transfer in building physics for more information.
It is possible that the heat flux sensor contributes significantly to
the total thermal resistance of the object (resistance error). In
this case the heat flux measurement must be corrected. For the
correction, see chapter 1. In order to limit the resistance error,
in all cases the contact between sensor and surrounding material
should be as well and as stable as possible, so that air gaps are
not influencing the measurement.
In a perfect environment, the initial calibration accuracy of heat
flux sensors is estimated to be +3 /- 3%.
In case of use of HFP01 on walls (insulating as well as bricks and
cements) the overall expected measurement accuracy for 12 hr
totals is +5 / -5 %.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 15/35
In case of analysis of thermal resistance of building envelopes,
the minimum recommended measurement time is 48 hours.
Hukseflux also offeres a complete measurement system for
analysis of building envelopes: TRSYS.
Figure 3.1 Estimation of convective, radiative and conductive
heat flux in building physics. The heat flux sensor is simply
mounted on or in the object of interest. This is typically in walls,
but can also be in the soil, e.g. on top of an underground heat
storage tank.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 16/35
4Specifications of HFP01
HFP01 is a heat flux sensor that measures the local heat flux
perpendicular to the sensor surface in the medium in which it is
incorporated or the object on which it is mounted. It can only be
used in combination with a suitable measurement and control
system. The HFP01 specifications except size, resistance,
sensitivity and weight are also applicable to sensor type HFP03;
see appendix.
HFP01 GENERAL SPECIFICATIONS
Specified
measurements Heat flux in W/m2perpendicular to the sensor surface
Installation See the product manual for
recommendations.
Temperature range -30 to +70 degrees C
Recommended number
of sensors Meteorological: two for each measurement station.
Building Physics: typically 1 or 2
sensors per measurement location
depending on building and wall
properties.
CE requirements HFP01 complies with CE directives
Series connection HFP01 sensors can be put electrically in
series to create a sensor with higher
sensitivity of better spatial resolution
using only one single readout channel.
The sensitivity then is the average of
the two sensitivities.
Thermal conductivity
dependence Eλ-0.07 % m.K/W (nominal value) λcal = 0
Temperature
dependence TD < +0.1%/ °C
Table 4.1 List of HFP01 specifications. (continued on next 2
pages)

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 17/35
HFP01 MEASUREMENT SPECIFICATIONS
Initial calibration
accuracy +3 /- 3%
Overall uncertainty
statement according to
ISO
estimated to be within +5 /- 5%, based
on a standard uncertainty multiplied by
a coverage factor of k = 2, providing a
level of confidence of 95%.
Application related errors should be
added to this error.
Expected typical
accuracy (12 hr totals)
of heat flux
measurement in soil
Initial calibration accuracy:
+3 /- 3%
Added errors within most common soils,
(clays, silts, organic), on most walls @
20 degrees C:
+0 / - 7%
Added typical temperature error: -10 +
40 degrees C
+2 / - 3%
Total typical value in soil (rounded off):
within
+5 / -15 %
Expected worst case
accuracy (12 hr totals)
of heat flux
measurement in soil
Initial calibration accuracy:
+3 /- 3%
Added errors with worst case soil, pure
sand @ 20 degrees C:
+0 / - 16%
Added worst case temperature error:
-30 + 70 degrees C
+5 / - 5%
Total worst case value in soil (rounded
off): within
+10 / -25 %
Expected typical
accuracy (12 hr totals)
of heat flux
measurement on a wall
Initial calibration accuracy:
+3 /- 3%
Added typical temperature error: -10 +
40 degrees C
+2 / - 3%
Total typical value on a wall (insulating,
brick, cement) (rounded off): within
+5 / -5 %
Low thermal resistance walls require
correction for the resistance error.
Table 4.1 List of HFP01 specifications. (started on previous page,
continued on next page)

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 18/35
HFP01 SENSOR SPECIFICATIONS
Esen (nominal) 50 µV/ W. m-2 (exact value on
calibration certificate)
λcal = 0, Tcal =20 °C
Sensor thermal
conductivity 0.8 W/mK
Sensor thermal
resistance Rth < 6.25 10 -3 Km2/W
Response time
(nominal) ± 3 min (equals average soil)
Range + 2000 to - 2000 W.m-2
Non stability < 1% change per year (normal
meteorological / building physics use)
Required readout 1 differential voltage channel
or possibly (less ideal)
1 single ended voltage channel.
When using more than one sensor and
having a lack of input channels, it can
be considered to put several sensors in
series, while working with the average
sensitivity.
Expected voltage
output Meteorology: -10 to - + 20 mV Building physics: -10 to 75 mV
(exposed to solar radiation)
Power required Zero (passive sensor)
Resistance 2 Ohm (nominal) plus cable resistance
Required programming ϕ= Vsen/ Esen
Sensor dimensions 80 mm diameter, 5 mm thickness
Cable length, diameter 5 meters, 5 mm
Weight including 5 m
cable, transport dim. 0.2 kg transport dimensions 32x23x3 cm
CALIBRATION
Calibration traceability to the “guarded hot plate” of National
Physical Laboratory (NPL) of the UK.
Applicable standards are ISO 8302 and
ASTM C177.
Recalibration interval Dependent on application, if possible
every 2 years, see appendix
OPTIONS
Extended cable Additional cable length x metres (add to
5m), AC100 amplifier, LI 18 hand held
readout, extended temperature range
Table 4.1 List of HFP01 specifications. (started on previous 2
pages)

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 19/35
5Short user guide
Preferably one should read the introduction and first chapters to
get familiarised with the heat flux measurement and the related
error sources.
The sensor should be installed following the directions of the
next paragraphs. Essentially this requires a data logger and
control system capable of readout of voltages, and capability to
perform division of the measurement by the sensitivity.
The first step that is described in paragraph 6 is and indoor test.
The purpose of this test is to see if the sensor works.
The second step is to make a final system set-up. This is
strongly application dependent, but it usually involves
permanent installation of the sensor and connection to the
measurement system.
Directions for this can be found in paragraphs 7 to 11.

Hukseflux Thermal Sensors
HFP01/ HFP03 manual version 0612 page 20/35
6Putting HFP01 into operation
It is recommended to test the sensor functionality by checking
the impedance of the sensor, and by checking if the sensor
works, according to the following table: (estimated time needed:
5 minutes)
Warning: during this part of the test,
please put the sensor in a thermally
quiet surrounding because a sensor
that generates a significant signal will
disturb the measurement.
Check the impedance of the sensor.
Use a multimeter at the 10 ohms
range. Measure at the sensor output
first with one polarity, than reverse
polarity. Take the average value.
The typical impedance of
the wiring is 0.1 ohm/m.
Typical impedance
should be 1.5 ohm for
the total resistance of
two wires (back and
forth) of each 5 meters,
plus the typical sensor
impedance of 2 ohms.
Infinite indicates a
broken circuit; zero
indicates a short circuit.
Check if the sensor reacts to heat flux.
Use a multimeter at the millivolt
range. Measure at the sensor output.
Generate a signal by touching the
thermopile hot joints (red side) with
your hand.
The thermopile should
react by generating a
millivolt output signal.
Table 6.1 Checking the functionality of the sensor. The procedure
offers a simple test to get a better feeling how HFP01 works, and
a check if the sensor is OK.
The programming of data loggers is the responsibility of the
user. Please contact the supplier to see if directions for use with
your system are available.