Think about modulation level as well as flow temperature. Most heatpumps at the start of a cycle seem to unnecessary use full output even when they should be able to predict it will result in a short cycle.
Every other thread seem to be people having problems with DWH, it must be possible to do better and make the setttings easier to understand.
How can a heatpump work for tenants who are unwilling to heat 24/7 and think a thermostat is a on/off switch while also automatically giving good COP for the next tenant who want 24/7 heating?
How will you cope when tenants don’t have a smart phone and will not pay for broadband? What the process when the tenant changes the wifi password on their broadband?
It not pratical to assume anyone will explain the system to tenants or hand over any instruction books! Letting agents just don’t do so, so the thermostat must explain everything that’s needed. Including when the pervious tenant had incorrectly changed many settings.
This is taken into account. There is both a thermal physics and COP model which the optimizer can exploit to find the best possible heating patterns.
If the thermostat is your main user interface then you can do both. In the end the system should adhere to the thermostat setpoint.
I currently don’t see a way to get a weather forecast into the heat pump without a wifi connection. In that case the system would simply run without a forecast, as the controller itself doesn’t depend on wifi. Robustness is key. You could add a mobile network chip as fallback, kind of like smart meters have already. But that also adds cost…
Agreed. And I think the systems we have today are already way too complicated. I want to have an auto mode that ‘just’ works without having to fiddle with it for hours and having to read all kinds of complicated manuals.
ps. for all my previous comments you have to make a clear distinction between the engineering side (internals) and what interfaces are exposed to the user (external)
I am familiar with it, interesting device! I think these kinds of devices already fall somewhat into the category of ‘energy management’. I think the heat pump should be controllable / integrate with as many external devices as possible to maximize interoperability. Many heat pumps are compatible with the SG-ready protocol, but it is too limited to interface with energy management devices that are currently being developed and will come on the market. In my personal opinion optimizing with respect to ToU tariffs is a function that an external device should impose as an optimization target on the heat pump, and not a function the heat pump should do on its own. (I hope that makes sense, energy management is an interesting topic as well)
I’m using a controller (thermostat) which uses both low load and weather compensation, with heating curve and PID that auto tunes itself based on the environment. It supports OpenTherm protocol and EMS protocol (EMS is now in beta). If you have smart TRVs, each can be used to determine the highest error (how far are we from the target temp) and user can choose whether to run in ECO = use primary temp sensor only or COMFORT = primary or secondary temp sensor(s) with the highest error. Forecasting is not even necessary since it monitors external and internal environment at all times, i.e. the target flow temperature is constantly adjusted.
I’m currently helping out with the development of low load control for boilers like mine that do not respect requests for 0% relative modulation. The feature is almost ready for wider audience.
The controller has capability to maintain room temperature within 0.1 C degrees from the target temperature. Here’s a screenshot of my overnight run last night.
But again, please keep in mind that I’m constantly testing various scenarios so the operations and graphs as not that stable yet as they will be once the development is completed.
Partly disagree, a heatpump should be easy to setup with a tariff that have a fixed cheap overnight period. For example it should automatically do DHW full reheat and finish the period after completing any pending defrost (and recovery). (Likewise with a single evening period with higher electricity prices.)
But Octopus Agile etc should not be built in as it depends on a API that will change over next 30 years. There is nothing that will make me belief a heatpump maker wil keep issuing updates for a 30 year old product…
Practically we have to move away from heatlose, radator sizing, and balancing being based on the usage of rooms at the time of the installation. For example it is common for a bedroom to change into a home office, or a living room to become a bedroom. Also if a person is ill, often require their bedroom to be heated to a higher temperature then livingroom with only a few hours notice of them being sent home from hospital.
Yet standard TRV results in many radiators being unnecessary completely removed from the system volume, and don’t workout how to maximise overall system flowrate along with minimizing flow temperature. (But it is unthinkable for each make of heatpump/controller to need a different type of TRV.)
There is the hard problem that the TRV user interface tend to result in people thinking they will save money by not heating some rooms, but with heatpumps it is often a bad ideal to have unheated rooms. So somehow the UI need to help the user understand, while not needing the user to have a smartphone or tablet.
@critictidier Nice project! Currently the heatpump can already do this, but using separate thermostats. This is also called a master-master setup, as opposed to a master-slave setup which has a single thermostat and dumb TRVs. But I think that setup is a lot less popular because it’s more complex and costs more to install.
Yes, this is basic PID control which is industry standard. I think we can do much better than that, so that’s what I am working on.
Agreed, it would be a nice feature. However more and more large energy consumers are starting to integrate these kinds of features, at some point you need a central unit to orchestrate it all. But I agree that some basic tariff based load shifting functionality would be nice though.
I am still working towards a renovation project which will include an Air Source Heat Pump - probably a Samsung Mono Pump with a matching climate hub. The Mono Pump combines the main ASHP outdoor components with things like the pump, expansion tank etc. and therefore saves on indoor space which in my case will be important.
I am not aware of what if any ASHP support OpenTherm control in the UK. Whilst the Netherlands has required all boiler manufacturers to support OpenTherm as an example, the UK does not. So whilst Vaillant and Worcester Bosch support OpenTherm on their boilers in the Netherlands, they do not even on the same models in the UK. (A convertor module is used to convert to OpenTherm and they do not include this in the UK and will regard use of the same module in the UK as invalidating their warranty.)
I have found several different 3rd party controllers for ASHP which all (of course) claim to work better than the built-in controllers. I have seen Homely as already mentioned here and it seems to have a good reputation and not only deliver better results but also supports smart energy tariffs.
For the benefit of everyone and further discussion here are the others I have identified. They are not all available in all countries and don’t all work with all ASHP brands. (At least yet.)
For completeness here is the link to Homely
Honeywell/Resideo claim that some of their smart Thermostats/Controllers are now ASHP compatible. Frustratingly their US and EU products are not necessarily equivalent and I am not sure I would trust their ASHP claims - yet.
As mentioned the Homely can integrate with smart energy tariffs - at least in the UK, the Tado product however has its own unique selling point in that it integrates with their own matching smart thermostats and smart TRVs.
If anyone has knowledge of any of these or others - please share.
I believe the Octopus Cosy 6 also has a lot of the components hidden within the outside unit, so might be worth a look.
Most UK heatpumps have their own proprietary connection, with a smart controller that is specific to that heatpump. What would be the advantage to having to buy additional 3rd party equipment when the heatpump should work optimally out of the box?
Nearly all heatpump standard controls depend on the installer/user understanding weather compensation curves and setting the corect curve.
A significant number of heatpumps sold in the UK don’t have load compensation, so can’t cope when the building heat lose is dependant on anything other then outside temperature.
Often the software is written by people who have never lived in a home with standard UK type radators.
Normal the software can’t be automatically updated
It very uncommon for the surport/programming team to be UK based and to remain working in same team after the product have shipped with a business model that depends on keeping existing users happy.
I don’t know of any heatpumps company that is commited to integrate with dynamic electricity tarrifs from multiple energy companies.
But some of the more costly heatpumps do have “good enough” combined weather/load compensation. However they cost more then a basic heatpump with Homely.
Yet Homely at present have no surport for reducing cycling of heatpumps, so is clearly not at good as build in auto optimising control systems could be.
The Octopus Cosy 6 have yet to be proven with independent monitoring, and don’t yet intergrate with other energy companies. But they seem to be heading in the correct direction.
The mitsubishi ecodan auto adapt seems to work well, but with quality issues on some of their flow/return sensors.
You make some good points @Ringi, though I still worry that a 3rd-party heating controller will ultimately conflict with whatever the heatpump wants to do, and getting the two to work in harmony will be quite difficult. (I know, I’ve tried to build my own). As you say, built-in weather/load compensation ought to be good enough, particularly if heatpumps are properly sized to begin with.
As for dynamic electricity tariffs, it doesn’t make sense to alter the heating in response to the grid price. In my experience, the heatpump reacts too slowly, and [my] property loses heat too quickly. Far better to use a battery to average out the unit cost, and let the heatpump coast along at a steady temperature.
In my opinion the heat pump should always be the master. It can accept external price signals and electrical / thermal constraints and then make a decision on what to do, but ultimately external controllers should never be able to access the same kind of functionality an internal controller has.
It depends a lot on the thermal characteristics of the house. A well-insulated house with an underfloor heating system will be able to optimize a lot more than someone with high temperature radiators and poor insulation. Once you add electricity prices, weather compensation, load compensation, building thermal characteristics etc. all together the complexity of the problem explodes. Hence why I tackled this in an optimization framework.
It appears that the indoor controller on the Ecodan (FTC) tells the outdoor unit what flow temperature it needs based on Fixed Flow / WC / Auto Adapt and then the outdoor unit controls itself to meet this.
I wonder if there is a better (although I assume much more complex) way to do this if you had a single controller getting the most out of the mechanics of the heat pump to meet the requirements of the building.
Ecodan Auto Adapt seems to be doing something to reduce cycling. So I expect it is being given information on modulation levels etc, so not just setting flow temperature based on hybred load/weather compensation.
Passiv is sold based on installers not needing to set weather compensation curve, is claims to use weather forecasts and to let owners get payments for taking part in “demand management services” that they organise and take a cut off.
It seems to use internal temperature, external temperature and weather forecasts letting owner set when they wish their home to be at a given temperature. So heatup in cheapest way to be warm on returning from work.
I think they are part of a large smart meter provider, so may be using same communication network.
Only sells to installers and large socal landlords etc.
They seem to directly control zone etc, but can’t find documents on their website. Eg it replaces some of the function of the heatpump wiring center. But use heatpump to control 3 port value for DHW.
I wasn’t so sure AA was as smart as that, I thought the reduction in cycling compared to WC was just that it allowed for larger over and undershooting of the target flow temperature and that it could monitor the indoor temperature to work out how much over or under shoot it could get away with before it started to have too much of an impact on comfort.
They maybe just setting flow temperature to a little over return temperature so heatpump starts cycle with low modulation, slowly increasing requested flow temperature based on return temperature.
Then when they detect end of cycle, set requested flow temperature to be low until they wish next cycle to start. With high thermal mass UFH such a system should be able to automatically detect that a heating cycle does not need to start until at least 30 minutes after the start of previous heating cycle.
Ah yes but that is exactly the point. Whilst yes they should work out of the box at full optimal settings - they don’t. Either they require a lot of tweaking as discussed in these forums or (allegedly) these 3rd party controllers especially the Homely can do a better job for you.
