EP2546589A1

EP2546589A1 - temperature control in a refrigerated transport container

Info

Publication number: EP2546589A1
Application number: EP11173525A
Authority: EP
Inventors: Leijn Johannes Sjerp Lukasse; Janneke Emmy de Kramer-Cuppen
Original assignee: AP Moller Maersk AS
Current assignee: AP Moller Maersk AS
Priority date: 2011-07-12
Filing date: 2011-07-12
Publication date: 2013-01-16

Abstract

Disclosed is a system for and a method of controlling temperature within a refrigerated transport container (1), the refrigerated transport container (1) comprising at least a transport volume (45), a control unit (7), and a cooling space (41), one or more evaporator fans (10) providing an air flow through the cooling space (41), where air passing through the cooling space passes at least a return air temperature sensor (5), a cooling unit (16), and a supply air temperature sensor (25), wherein the method comprises controlling unmeasured temperatures in the transport volume (45) within a temperature range adjacent to a setpoint or target temperature (Tset), using two or more transport volume temperature indicators, where the indicators are based on at least measured supply air temperature and/or measured return air temperature.

In this way, control of unmeasured temperatures in the transport volume is provided that enables improved control over temperatures of the loaded perishable produce thereby reducing the rate of quality loss of the transported produce.

Description

Disclosed is a method of and a system for controlling temperature within a refrigerated transport container, or other refrigerated storage spaces.

BACKGROUND

Temperature in a refrigerated transport container, or another kind of refrigerated storage space, is typically controlled within a temperature range adjacent to a setpoint or target temperature (forth referred to as setpoint temperature or setpoint). The refrigerated transport container may for example comprise an insulated enclosure divided in a cooling space and a transport volume. Typically, the transport volume is loaded with perishable produce such as meat, vegetables and fruit, etc. The setpoint temperature is then typically chosen to reduce quality degradation of the perishable produce.
The cooling space may e.g. be separated from the transport volume by a panel equipped with one or more openings to allow a return air flow from the transport volume into the cooling space and a supply air temperature flow from the cooling space into the transport volume.
The air flow through the cooling space typically passes at least a return air temperature sensor, a device for reducing the temperature of the passing air, e.g. a cooling unit or system, and a supply air temperature sensor. In such systems, the return air temperature sensor typically measures the temperature of air returning from the transport volume while the supply air temperature sensor measures the temperature of air supplied to the transport volume.
Temperature control protocols may selectively control a cooling unit coupled to the refrigerated transport container in order to maintain the setpoint temperature inside the refrigerated transport container.
One typical type of a cooling unit or refrigeration unit used in refrigerated storage transport containers is based on the so-called vapour compression refrigeration cycle. This cycle comprises at least a compressor, a condenser, an expansion device, an evaporator and a capacity regulating device. The compressor sucks refrigerant vapour from the evaporator and compresses the refrigerant vapour which subsequently flows to the condenser at high pressure. The condenser ejects its heat to a medium outside the refrigerated transport container while condensing the refrigerant vapour. The liquefied refrigerant then flows to the expansion device in which a refrigerant pressure drops. The low pressure refrigerant then flows to the evaporator where the refrigerant evaporates while extracting the required heat from the refrigerated transport container.
Other typical cooling units or refrigeration units used in refrigerated transport containers may be different.
Temperatures in the transport volume are typically unmeasured. In a steady state operation, measured supply air temperature may normally be a fairly accurate representative of a coldest temperature in the transport volume. In the steady state operation, measured return air temperature may usually be a reasonable representative of average temperature in the transport volume. In the steady state operation, a warmest temperature in the transport volume is usually a little higher than return air temperature, but remains unknown and e.g. depends on the way the cargo is stowed inside the container.
For frozen commodities, typically shipped at setpoints below -10 °C and usually around -20 °C, it is especially important that produce temperature is not too far above setpoint. Therefore for setpoints below -10 °C, it is common practice to control a measured return air temperature closely to the setpoint.
For chilled commodities, typically shipped at setpoints above -10 °C, both too high and too low produce temperatures are undesirable. The adverse effect of too high above setpoint is fairly obvious; that is the whole reason why refrigeration is applied. However being too low below setpoint, chilled commodities may actually suffer as well. Some chilled commodities are susceptible to chilling injury, e.g. like bananas turning grey in home fridges.
Furthermore, many chilled commodities are susceptible to freezing injury, which especially becomes an issue when sensitive commodities like grapes are shipped at setpoints just above their freezing point.
Traditionally, refrigerated transport containers used to be stuffed with produce which was already pre-cooled to a temperature close to setpoint, so transport volume temperatures were always more or less in the steady state condition.
The current practice however, is that ever more containers are stuffed with warm produce right after harvest, whereby it is up to the container's cooling unit to reduce produce temperature from stuffing temperature to a temperature range adjacent to the setpoint temperature. In the banana trade for example, it is now standard operations procedure to load uncooled bananas of around 25 °C in containers operating at a setpoint of about 13.5 °C. In these non-steady state conditions, return air temperature becomes a poor indicator of the warmest temperature inside the transport volume.
Typically, the warmest temperature converges a lot slower to a temperature range adjacent to a setpoint temperature than return air temperature.
In view of the increasing number of warmly-stuffed containers, there is a need to effectively and efficiently manipulate measured supply and return air temperature in order to ensure that actual transport volume temperatures reside as much as possible and as quickly as possible within a desired temperature range adjacent to a setpoint temperature.

SUMMARY

It is an object to provide a temperature control for a refrigerated transport container more advanced than just controlling either return air temperature or supply air temperature to a setpoint. The temperature control ensures that a larger portion of the transport volume temperatures resides in a desired temperature range adjacent to a setpoint temperature during a larger part of the transport time.
A first aspect relates to a method of controlling temperature within a refrigerated transport container, the refrigerated transport container comprising at least a transport volume, a control unit, and a cooling space, one or more evaporator fans providing an air flow through the cooling space, where air passing through the cooling space passes at least a return air temperature sensor, a cooling unit, and a supply air temperature sensor, wherein the method comprises:

controlling unmeasured temperatures in the transport volume within a temperature range adjacent to a setpoint or target temperature, using two or more transport volume temperature indicators, where the indicators are based on at least measured supply air temperature and/or measured return air temperature.

Average produce temperature within the refrigerated transport volume typically lies somewhere in-between the supply air temperature and a few degrees above the return air temperature due to temperature gradients within the transport volume.
An advantage of controlling unmeasured temperatures in the transport volume, instead of just supply or return air temperature, within a temperature range adjacent to a setpoint or target temperature (Tset), is that this improves control over temperatures of the loaded perishable produce.
The motivation for transporting perishable commodities in refrigerated transport containers is that their quality loss depends on temperature. Moreover, the rate of quality loss deteriorates at suboptimal temperatures.
Controlling temperatures in the transport volume helps to reduce the rate of quality loss. Especially in pulldown situations, occurring in warmly-stuffed containers, the advantage may be significant because then the difference between produce temperature and either supply or return air temperature is largest.
In one embodiment, the at least two transport volume temperature indicators are one or more selected from the group consisting of:

current and/or recent supply air temperature, or a function thereof,
current and/or recent return air temperature, or a function thereof,
an estimator for temperature in a coldest spot of the transport volume,
one or more estimators for temperatures in one or more warmer spots in the transport volume,

current and/or recent return air temperatures Tret and/or,
current and/or recent supply air temperatures Tsup and/or,
earlier estimates if available and/or
a history of power supply to the cooling unit within a predetermined period of time (e.g. the last 24 hours or so).

Temperatures in the transport volume are unmeasured and therefore cannot be controlled directly. The use of transport volume temperature indicators, correlated to temperatures in the transport volume, advantageously enable indirect control over temperatures in the transport volume, more than just controlling return or supply air temperature to a setpoint.
The estimators may e.g. be initialized or re-initialized after a power cut or powering down based on the latest estimate made just before the power cut or power down happened e.g. taking into account the duration of the power cut. One example may e.g. be that the initial estimate after power is established again is equal to the estimate at the power cut or power down plus a factor (e.g. 0.1 °C/h) times the duration of the period (h) of time without power.
In one embodiment, the estimator for temperature in a coldest spot of the transport volume estimates temperature in a coldest spot of the transport volume based on current and/or recent supply air temperatures and one or more previous estimates of the temperature in a coldest spot of the transport volume, and/or the one or more estimators for temperatures in one or more warmer spots in the transport volume estimates temperatures in one or more warmer spots of the transport volume based on current and/or recent supply air temperatures, current and/or recent return air temperatures, and one or more previous estimates for temperatures in one or more warmer spots in the transport volume.
The estimator for temperature in a coldest spot of the transport volume may e.g. be an estimator whose change is based on a function of current and/or recent supply air temperatures and one or more previous estimates of the temperature in a coldest spot of the transport volume.
The estimator for temperatures in one or more warmer spots in the transport volume may e.g. be an estimator whose change is based on a function of the current and/or recent supply air temperatures, current and/or recent return air temperatures, and one or more previous estimates for temperatures in one or more warmer spots in the transport volume.
When to-be-controlled states of any dynamic process are unmeasured, the use of estimators for that states advantageously offer the possibility to have some degree of control over those states. Temperatures in the transport volume are unmeasured, yet some degree of control becomes possible by using estimators for temperature (Tcold) in a coldest spot of the transport volume and one or more estimators for temperatures (Twarm) in one or more warmer spots in the transport volume. The estimators could for example be mathematical filters mapping available information on current and/or recent supply air temperature and current and/or recent power supply to the rate of temperature change at the coldest and one or more warmer locations in the transport volume. These filters could be tuned using earlier collected experimental measurements of trajectories of supply air temperature and temperature in the coldest and one or more warmer locations in the transport volume.
In one embodiment, the method comprises:

using an estimator for temperature in a coldest spot of the transport volume and one or more estimators for temperatures in one or more warmer spots of the transport volume, and controlling a weighted average of these estimators to the temperature setpoint (e.g. plus an offset, where the offset maybe zero).

Controlling a weighted average of an estimate for temperature (Tcold) in a coldest spot of the transport volume and one or more estimators for temperatures (Twarm) in one or more warmer spots of the transport volume offers an important advantage over just controlling supply or return air temperature to setpoint: it controls a true representative of produce temperature to setpoint.
In one embodiment, the method comprises:

constraining the estimator for temperature in the coldest spot to a minimum constraint and/or a maximum constraint.

Including maximum and minimum constraints advantageously helps to avoid the exceeding of temperature limits that are critical to produce quality. Especially important are the limits in chilled mode below which chilling injury or freezing injury may be inflicted, or the limit in frozen mode above which the carried commodity may start to thaw. A well-known example of chilling injury is the dull grey coloration of bananas stored in home fridges. The risk of freezing injury especially exists for all fruit stored at temperatures just above their freezing point (for example the pale brown coloration of grapes and their stems).
In one embodiment, the method comprises:

using supply air temperature or a time-averaged function thereof, and return air temperature or a time-averaged function thereof as indicators for the coldest and the warmest temperature in the transport volume, respectively, and
controlling a weighted average of the supply air temperature and the return air temperature to the temperature setpoint entered into the control unit (e.g. plus an offset, where the offset maybe zero).

In calculation of the weighted average, the weight of supply air temperature may differ from the weight of the return air temperature.
Controlling a weighted average of an estimate for temperature (Tcold) in a coldest spot and an estimate for temperature (Twarm) in a warmest spot of the transport volume offers an important advantage over just controlling supply or return air temperature to setpoint: it controls a true representative of produce temperature to setpoint. Supply air temperature (Tsup) or a time-averaged function thereof, and return air temperature (Tret) or a time-averaged function thereof, are not the most advanced estimators for the coldest and the warmest temperature in the transport volume, but the advantage is that they are straightforwardly available in any refrigerated transport container.
In one embodiment, the method comprises:

constraining the supply air temperature, or a time-averaged function thereof, to a minimum constraint and/or a maximum constraint.

Including maximum and minimum constraints advantageously help to avoid the exceeding of temperature limits that are critical to produce quality, as explained above.
In one embodiment, the method comprises:

controlling by a slave-controller the supply air temperature or a time-averaged function thereof to a supply air temperature setpoint, and adjusting the supply air temperature setpoint as a function of a temperature setpoint and a measured return air temperature by a master-controller.

This is to some extent an alternative implementation of the embodiment controlling a weighted average of the supply air temperature and the return air temperature to the temperature setpoint with similar advantages.
An additional advantage of using the master-slave concept is the possibility to use the master controller to make the supply air temperature setpoint any possible function of current and/or recently measured return air temperature and to also shape the dynamics of the response of supply air temperature to changes in return air temperature.
In one embodiment, the master-controller adjusts the supply air temperature setpoint such that the weighted average of the supply air temperature and the return air temperature substantially equals the temperature setpoint (e.g. plus an offset, where the offset maybe zero).
In calculation of the weighted average the weight of supply air temperature may differ from the weight of the return air temperature.
This advantageously combines the advantages provided by the master-slave concept as used in the preceding embodiment with the advantage of controlling a weighted average of an easily available estimate for temperature (Tcold) in a coldest spot and an easily available estimate for temperature (Twarm) in one or more warmer spots of the transport volume, which is the control of a true representative of produce temperature to setpoint.
In one embodiment, the method comprises

constraining the supply air temperature setpoint, as adjusted by the master-controller, to a minimum constraint and/or a maximum constraint.

Including maximum and minimum constraints advantageously helps to avoid the exceeding of temperature limits that are critical to produce quality, as explained above.
In one embodiment, the value for the minimum constraint and/or the maximum constraint is dependent on the temperature setpoint and/or the time elapsed since activation of the controller.
Making maximum and minimum constraints dependent on the temperature setpoint and/or the time elapsed since activation of the controller advantageously increases flexibility to tailor the constraints to the actual need. At for example a setpoint of -20 °C a maximum constraint should be close to setpoint, because for frozen commodities it is only important that produce temperatures stay below a certain level. At for example a setpoint of 0 °C a minimum constraint should be close to setpoint to avoid freezing injury, while a maximum constraint might be more tolerant. The time elapsed since activation of the controller correlates to lowest temperature in the transport volume. Therefore for example in a warmly-stuffed container with grapes right after activation of the controller at power-up a supply air temperature multiple degrees C below the freezing point will not freeze the grapes, while later on that risk increases. So a minimum constraint tightening over time may be appropriate.
In one embodiment, the refrigerated transport container is not a transport container but another type of refrigerated space in connection with a cooling unit. This could for example be an item of refrigerated road transport equipment, a reefer ship, or any type of stationary cold storage room.
A second aspect relates to a system for controlling temperature within a refrigerated transport container, the refrigerated transport container comprising at least a transport volume, and a cooling space, one or more evaporator fans providing an air flow through the cooling space, where air passing through the cooling space passes at least a return air temperature sensor, a cooling unit, and a supply air temperature sensor, wherein the system comprises a control unit adapted to:

control unmeasured temperatures in the transport volume within a temperature range adjacent to a setpoint or target temperature, using two or more transport volume temperature indicators, where the indicators are based on at least measured supply air temperature and/or measured return air temperature.

The embodiments of the system correspond to the embodiments of the method and have the same advantages for the same reasons.

BRIEF DESCRIPTION OF THE DRAWINGS

Preferred embodiments of the invention will be described in more detail in connection with the appended drawings, in which:

Figure 1 schematically illustrates a simplified longitudinal cross-sectional view of a refrigerated space in the form of a refrigerated transport container;
Figure 2 schematically illustrates a block diagram representing a so-called master-slave controller according to one embodiment;
Figure 3 presents a computer simulation output schematically illustrating a setpoint (Tset) entered into a controller and temperature trajectories for a temperature of the supply air flow (Tsup), a temperature of the return air flow (Tret) and a warmest produce temperature (Twarm) in the transport volume in a situation where Tsup is controlled to the entered Tset;
Figure 4 presents another computer simulation output schematically illustrating a setpoint (Tset) entered into a master-controller and temperature trajectories for a temperature of the supply air flow (Tsup), a temperature of the return air flow (Tret), a warmest produce temperature (Twarm), and a slave-controller's setpoint (Tset-slave) adjusted by a master controller;
Figure 5 schematically illustrates measurements collected in a real transport container where temperature is controlled like in Figure 3;
Figure 6 schematically illustrates measurements collected in a real transport container where temperature is controlled like in Figure 4.

DESCRIPTION OF PREFERRED EMBODIMENTS

Figure 1 schematically illustrates a simplified longitudinal cross-sectional view of a refrigerated space in the form of a refrigerated transport container.
Shown is one example of a refrigerated transport container 1, or another type of refrigerated storage space, comprising at least a transport volume 45, a control unit 7, and a cooling space 41. The cooling space 41 may be situated inside an insulated enclosure of the transport container 1 and may (as shown) be separated from the transport volume 45 by a panel or the like equipped with one or more openings to allow a return air flow 50 into the cooling space 41 and a supply air flow 55 out of the cooling space 41.
The air flow through the cooling space may be maintained by for example one or more evaporator fans 10 or one or more other units providing a similar function. On its way through the cooling space 41, air successively passes at least a return air temperature sensor 5, the one or more evaporator fans 10, a cooling unit or system 16 (or one or more other units with a similar function) reducing the temperature of the passing air, and a supply air temperature sensor 25.
In this kind of system, the return air temperature sensor 5 measures the temperature of air returning from the transport volume (forth denoted Tret), while the supply air temperature sensor 25 measures the temperature of air supplied to the transport volume (forth denoted Tsup).
Unmeasured temperatures in the transport volume (45) are controlled by the controller (7) to be within a temperature range adjacent to a setpoint temperature (Tset) using two or more transport volume temperature indicators, where the indicators are based on at least measured supply air temperature and/or measured return air temperature. As a result the temperature control is more advanced than just controlling supply or return air temperature to a setpoint Tset, like in traditional chilled respectively frozen mode operation. For example the average temperature of the supply air temperature Tsup may temporarily be allowed to be below the setpoint Tset in order to speed up the pulldown of procude temperatures in the transport volume.
The controller (7) may e.g. comprise a master-slave controller setup as explained in connection with Figure 2 or its functionality could be provided in another fashion.
Further aspects and variations will be explained further in the following.
Figure 2 schematically illustrates a block diagram representing a so-called master-slave controller according to one embodiment. In this embodiment, the process 217 represents temperature dynamics within a refrigerated transport container (see e.g. 1 in Figure 1). Though each location in the refrigerated transport container has its own temperature 219, only two of them are measured: a Return air Temperature Sensor 5 measures the return air temperature Tret 213 and a Supply air Temperature Sensor 25 measures the supply air temperature Tsup 209.
This block diagram represents a so-called master-slave controller 200 according to one embodiment where an entered setpoint Tset 201 generally is first processed in a master controller 203 that based on Tset 201 and Tret 213 manipulates or derives a second or modified setpoint Tset_slave 205. The difference between the modified setpoint Tset_slave 205 and supply air temperature Tsup 209 is then received by the slave controller 207, which then aims to minimize this difference, effectively controllingTsup 209 to the modified setpoint Tset-slave 205 by adjusting the amount of heat absorbed by the cooling unit (see e.g. 16 in Figure 1) in a cooling space of the refrigerated transport container, which in this schematic representation may be regarded to be part of the process 217.
In the present embodiment, the user's setpoint Tset 201 is treated as a setpoint to a master controller 203 where the master controller 203 manipulates the slave setpoint Tset_slave 205. The slave controller 207 then controls the supply air temperature Tsup 209 to the slave setpoint Tset_slave 205. The slave setpoint Tset-slave 205 deliberately deviates from the master setpoint Tset 201 with the objective to control the average of Tsup 209 and Tret 213 to the setpoint Tset 201. By allowing the average Tsup 209 to be below Tset 201 instead of controlling it to Tset, a larger portion of the temperatures 219, including produce temperatures, in the container will be in a temperature range adjacent to setpoint Tset 201 and will be so quicker.
When a controller (see e.g. 7 in Figure 1) initiates, for example when the unit powers up, Tset-slave 205 may be initialized as a function of Tset 201 and Tret 213, for example according to Tset_slave = Tset - 0.5x(Tret-Tset). This lowers, in this specific example, the modified or effective supply air temperature setpoint with half the difference between the temperature of the return air and the normal setpoint. It is to be understood that other suitable initializations may be used. What is significant is that the modified or effective supply air temperature setpoint Tset-slave is lowered initially in proportion to the difference between return air temperature and setpoint Tset.
Following this initialization Tset_slave 205 may then be updated by the master controller 203 at the beginning of each subsequent cycle e.g. according to: $Tset_slave (k + 1) =$
$\max (Tset_slave_min; (1 - 0.2 \times tcycle / 60) \times Tset_slave (k) + 0.2 \times tcycle / 60 \times (2 \times {Tset -}^{\overline{T} ret} (k))) [°C],$

where
k designates the k-th cycle,
tcycle = duration of the preceding cycle [minutes],
T _rgt (k) = return air temperature averaged over the k-th cycle [°C], Tset_slave(k) = slave setpoint during the k-th cycle, and
Tset-slave-min = a lower constraint on Tset-slave, meant to avoid freezing or chilling injury and e.g. given by Tset_slave = Tset - 1 °C.
In the equation above, a cycle is a predefined period of time, which may be constant or may be defined otherwise. For example, in systems with on/off controlled compressors it may be defined as a period of time from one start of a compressor until its next start.
The preceding equation helps to control the average of Tsup and Tret to Tset. This can be seen by observing that a control objective 'average of Tsup and Tret = Tset' is equivalent to ' ` (Tsup+Tret)/2 = Tset' is equivalent to 'Tsup = 2xTset-Tret'. If we assume that Tsup = Tset-slave, something the slave-controller may take care of, then 'Tsup = 2xTset-Tret' is equivalent to the control objective 'Tset_slave = 2xTset-Tret'. A very simple implementation of this, is to program the master controller according to: $Tset_slave (k + 1) = \max (Tset_slave_min; 2 \times {Tset -}^{\overline{T} ret} (k)) [°C]$
However, any high-frequent fluctuation in ^Tret (k) is just passed on to Tset_slave(k+1). This could then result in undesired high-frequent oscillations in Tset_slave. To avoid this behaviour, a low pass filter is added. One example of a simple low-pass filter is a linear difference equation of the type Tset_Slave(k+1) = (1- smoothing factor) x Tset_slave(k) + smoothing factor x ^Tret (k), which is used in the preceding paragraph, using a 'smoothing factor = 0.2xtcycle/60'.
Figure 3 schematically illustrates a computer simulation with a setpoint (Tset) 301 entered into a controller and temperature trajectories for a temperature of the supply air flow (Tsup) 302, a temperature of the return air flow (Tret) 303 and a warmest produce temperature (Twarm) 304 in the transport volume.
In this situation Tsup 302 is controlled to the entered Tset 301. This reflects a traditional approach to temperature control in chilled mode operation. It could be achieved by a control set-up as depicted in Figure 2 where the master controller just sets Tset_slave to Tset 301, although a more natural implementation would then be to omit the master controller and just feed the difference between Tset 301 and Tsup 302 to the slave controller (which then in effect becomes a master controller or the only controller for this purpose).
In traditional frozen mode operation, Tret 303 would be controlled to Tset 301. In that situation, the temperature pulldown would proceed at maximum cooling capacity until the curve of Tret 303 reaches setpoint, regardless how much Tsup 302 undershoots the setpoint Tset 301.
Figure 3 illustrates the traditional approach in chilled mode operation, i.e. operation at setpoints above -10 °C. In real shipments the warmest produce temperature Twarm 304 in the transport volume is normally unmeasured, but the computer simulation shows a realistic pattern.
Figure 4 shows a computer simulation with simulated trajectories for temperature Tsup 302, Tret 303, Twarm 304 resulting from entering the setpoint Tset 301 into a master-controller, which then manipulates the slave-controller's setpoint Tset_slave 305. The slave-controller's setpoint Tset_slave 305 is adjusted by the master controller, that based on Tset 301 and Tret 303 manipulates the setpoint Tset_slave 305 (constrained to Tset_slave ≥ Tset -1) with the objective to control the average of Tsup 302 and Tret 303 to Tset 301, while the slave controller aims to minimize the difference between supply air temperature Tsup 302 and its adjusted supply air temperature setpoint Tset_slave 305 .
This master-slave controller is an implementation of the embodiment depicted in Figure 2 with the master-controller executing the algorithm as described in relation to Figure 2.
Comparing Figure 3 and Figure 4 illustrates that a faster temperature pulldown, i.e. a faster approach of the temperature to the setpoint, is achieved due to the master-slave control in Figure 4, while yet maintaining control over Tsup 302. For example after 2 days in Figure 3, Twarm 304 is still 6.7 °C, while in Figure 4 Twarm 304 then is already down to 6 °C. This is achieved by allowing supply air temperatures Tsup 302 colder than Tset 301. In general this means an increased risk of chilling injury. However the period of coldest Tsup 302 typically occurs in the beginning of the pulldown when temperatures in most locations in the transport volume are still above Tset 301. Consequentially the risk of inducing chilling injury is very limited while the benefit of faster pulldown is clear, namely less quality degradation due to too high temperatures (i.e. the whole idea of applying refrigeration).
In frozen mode operation the master-slave concept may be used for example to limit the undershoot of Tsup 302 during temperature pulldown like in Figure 4. This would for example offer the advantage of some energy saving at the expense of a slightly slower pulldown of warmest temperature Twarm 304 in the transport volume.
Figure 5 and Figure 6 show the trajectories of Tsup 302 and Tret 303 registered during two test shipments. It concerns two refrigerated transport containers making the same journey simultaneously. The containers both carry a cargo of warmly-stuffed citrus. The high initial cargo temperature causes high return air temperatures during the initial days of the voyage.
Figure 5 shows the trajectories of Tsup 302 and Tret 303 registered in a container where Tsup 302 is controlled to Tset 301, like in the simulation in Figure 3. Note that the persistent 0.2 °C offset between Tsup 302 and Tset 303 in Figure 5 is a consequence of a difference between the supply air temperature recorder sensor used to record the temperature measurements and the supply air temperature controller sensor (not shown; see e.g. 5 in Figure 1).
Figure 5 schematically illustrates a setpoint Tset 301 entered into a controller and temperature trajectories for a temperature of the supply air flow Tsup 302, and a temperature of the return air flow Tret 303. Like in Figure 3, the supply air temperature Tsup 302 is controlled to the entered Tset 301. Figure 5 does not contain the warmest produce temperature Twarm, as e.g. shown in Figure 3, as in real shipments this is unknown.
Figure 6 displays the recorded Tsup 302 and Tret 303 in a container controlled according to the concept shown in Figure 2 and simulated in Figure 4. It schematically illustrates a setpoint Tset 301 entered into a controller and temperature trajectories for a temperature of the supply air flow Tsup 302, and a temperature of the return air flow Tret 303. Figure 6 does not contain the warmest produce temperature Twarm as this is not known in real shipments.
Figure 6 illustrates how the master controller, deriving Tset_slave, e.g. as described in connection with Figure 2, responds to the high initial Tret 303 by reducing Tset_slave (not shown, but approximately equal to Tsup 302) to its lower bound Tset 301 minus 1 °C. Consequentially the pulldown of Tret 303 is faster. Later on, Tret 303 comes ever closer to Tset 301, while the master controller gradually rises Tset_slave with the obective to control the average of Tsup 302 and Tret 303 to Tset 301.
In Figure 6, a minor jitter is observable on Tsup 302. This is caused by the on/off control method implemented in the slave controller, with the excitations of Tsup 302 smoothened again to a large extent by displaying hourly averaged values of Tsup 302 in Figure 6.
In both charts (Figure 5 and Figure 6) the rise of Tsup 302 and Tret 303 up to 8 - 9 °C around 091220-00 is typically the result of a few hours without electric power supply, during which the container was moved from land to ship. Other power off periods due to unknown factors occur in Figure 5 around 091228-12 and in Figure 6 around 091224-00. Also in both charts. minor spikes of about 1 °C are visible in Tret 303 in a regular frequency. These are due to so-called defrosts, marked on the horizontal axis with little cubes 306 on the horizontal axis. In Figure 5 these occur more or less once a day, in Figure 6 less frequent. During a defrost period, a defrost control algorithm, e.g. implemented in the same control unit (7 in Figure 1), overrules the temperature controller, stops cooling, stops the evaporator fans (10 in Figure 1) and supplies heat to the cooling unit (16 in Figure 1) in order to remove frost formed on the cooling unit. Once the defrost controller terminates the defrost, the evaporator fans resume the air circulation and the temperature controller resumes temperature control.

Claims

A method of controlling temperature within a refrigerated transport container (1), the refrigerated transport container (1) comprising at least a transport volume (45), a control unit (7), and a cooling space (41), one or more evaporator fans (10) providing an air flow through the cooling space (41), where air passing through the cooling space passes at least a return air temperature sensor (5), a cooling unit (16), and a supply air temperature sensor (25), wherein the method comprises:
- controlling unmeasured temperatures in the transport volume (45) within a temperature range adjacent to a setpoint or target temperature (Tset), using two or more transport volume temperature indicators, where the indicators are based on at least measured supply air temperature and/or measured return air temperature.
The method according to claim 1, wherein the at least two transport volume temperature indicators are one or more selected from the group consisting of:
- current and/or recent supply air temperature (Tsup), or a function thereof,

- current and/or recent return air temperature (Tret), or a function thereof,

- an estimator for temperature (Tcold) in a coldest spot of the transport volume (45),

- one or more estimators for temperatures (Twarm) in one or more warmer spots in the transport volume (45),
where upon activation of the controller (7), the estimators are initialized using:
- current and/or recent return air temperatures Tret and/or,

- current and/or recent supply air temperatures Tsup and/or,

- earlier estimates if available and/or

- a history of power supply to the cooling unit within a predetermined period of time.
The method according to claim 2, wherein
- the estimator for temperature (Tcold) in a coldest spot of the transport volume (45) estimates temperature (Tcold) in a coldest spot of the transport volume (45) based on current and/or recent supply air temperatures (Tsup) and one or more previous estimates of the temperature (Tcold) in a coldest spot of the transport volume (45), and/or

- the one or more estimators for temperatures (Twarm) in one or more warmer spots in the transport volume (45) estimates temperatures (Twarm) in one or more warmer spots of the transport volume (45) based on current and/or recent supply air temperatures (Tsup), current and/or recent return air temperatures (Tret), and one or more previous estimates for temperatures (Twarm) in one or more warmer spots in the transport volume (45).
The method according to any one of claims 1 - 3, wherein the method comprises:
- using an estimator for temperature (Tcold) in a coldest spot of the transport volume and one or more estimators for temperatures (Twarm) in one or more warmer spots of the transport volume (45), and controlling a weighted average of these estimators to the temperature setpoint (Tset).
The method according to any one of claims 1 - 2, wherein the method comprises:
- using supply air temperature (Tsup) or a time-averaged function thereof, and return air temperature (Tret) or a time-averaged function thereof as indicators for the coldest and the warmest temperature in the transport volume, respectively, and

- controlling a weighted average of the supply air temperature and the return air temperature to the temperature setpoint entered into the control unit.
The method according to any one of claims 1 - 5, wherein the method comprises:
- controlling by a slave-controller the supply air temperature or a time-averaged function thereof to a supply air temperature setpoint (Tset_slave), and adjusting the supply air temperature setpoint (Tset_slave) as a function of a temperature setpoint (Tset) and a measured return air temperature by a master-controller (203).
The method according to claim 6, wherein the adjustment of the supply air temperature setpoint is made such that the weighted average of the supply air temperature and the return air temperature substantially equals the temperature setpoint (Tset).
The method according to any of the preceding claims where the refrigerated transport container is not a transport container but another type of refrigerated space in connection with a cooling unit.
A system for controlling temperature within a refrigerated transport container (1), the refrigerated transport container (1) comprising at least a transport volume (45), and a cooling space (41), one or more evaporator fans (10) providing an air flow through the cooling space (41), where air passing through the cooling space passes at least a return air temperature sensor (5), a cooling unit (16), and a supply air temperature sensor (25), wherein the system comprises a control unit (7) adapted to:
- control unmeasured temperatures in the transport volume (45) within a temperature range adjacent to a setpoint or target temperature (Tset), using two or more transport volume temperature indicators, where the indicators are based on at least measured supply air temperature and/or measured return air temperature.
The system according to claim 9, wherein the at least two transport volume temperature indicators are one or more selected from the group consisting of:
- current and/or recent supply air temperature (Tsup), or a function thereof,

- current and/or recent return air temperature (Tret), or a function thereof,

- an estimator for temperature (Tcold) in a coldest spot of the transport volume (45),

- one or more estimators for temperatures (Twarm) in one or more warmer spots in the transport volume (45),
where upon activation of the controller (7), the estimators are initialized using:
- current and/or recent return air temperatures Tret and/or,

- current and/or recent supply air temperatures Tsup and/or,

- earlier estimates if available and/or

- a history of power supply to the cooling unit within a predetermined period of time.
The system according to claim 10, wherein
- the estimator for temperature (Tcold) in a coldest spot of the transport volume (45) estimates temperature (Tcold) in a coldest spot of the transport volume (45) based on current and/or recent supply air temperatures (Tsup) and one or more previous estimates of the temperature (Tcold) in a coldest spot of the transport volume (45), and/or

- the one or more estimators for temperatures (Twarm) in one or more warmer spots in the transport volume (45) estimates temperatures (Twarm) in one or more warmer spots of the transport volume (45) based on current and/or recent supply air temperatures (Tsup), current and/or recent return air temperatures (Tret), and one or more previous estimates for temperatures (Twarm) in one or more warmer spots in the transport volume (45).
The system according to any one of claims 9 -11, wherein the controller (7) is adapted to:
- use an estimator for temperature (Tcold) in a coldest spot of the transport volume and one or more estimators for temperatures (Twarm) in one or more warmer spots of the transport volume (45), and controlling a weighted average of these estimators to the temperature setpoint (Tset).
The system according to any one of claims 9 - 10, wherein the controller (7) is adapted to:
- use supply air temperature (Tsup) or a time-averaged function thereof, and return air temperature (Tret) or a time-averaged function thereof as indicators for the coldest and the warmest temperature in the transport volume, respectively, and

- control a weighted average of the supply air temperature and the return air temperature to the temperature setpoint entered into the control unit.
The system according to any one of claims 9 - 13, wherein the controller (7) is adapted to:
- control by a slave-controller the supply air temperature or a time-averaged function thereof to a supply air temperature setpoint (Tset_slave), and adjust the supply air temperature setpoint (Tset_slave) as a function of a temperature setpoint (Tset) and a measured return air temperature by a master-controller (203).
The system according to claim 14, wherein the adjustment of the supply air temperature setpoint is made such that the weighted average of the supply air temperature and the return air temperature substantially equals the temperature setpoint (Tset).