Network Thermal Storage vs Dedicated Thermal Storage

When it comes to district heating I’ve often heard it said that there’s no point in adding a dedicated thermal storage vessel because there’s already plenty of storage in the network.
I just tested this theory on a real project. In this case the volume of stored water in the network was calculated to be 6m3. The principle in trying to use the volume of the network as a store is that you operate a flow temperature deadband of say 5C. So when the flow reaches 75C you turn off the supply plant and let the flow temperature drop to 70 before switching the supply plant back on.
Assuming this deadband of 5C then this water volume represents roughly 0.03MWh of energy storage in the pipes. So for a, say, 2MW heat pump running at full output you can run for 54 seconds before you have to switch off. In those 54seconds you’ve raised the 6m3 back up to 75C. In reality it’s little more complicated than that because if during those 54seconds there’s a demand on the network the demand will be sucking heat out of the stored water simultaneously to the heat pump putting energy in so it’s a little longer depending on what the demand is. If the demand was an average of 1MW it would take a further 54seconds to fill the store.
Once the heat pump has switched off then the network flow temperature will drop down the deadband and the heat pump will switch back on at 70C. If the load during this off period were 2MW it would take 54 seconds to reach 70C and trigger the heat pump to switch on – if 1MW then 108 seconds, if 200kW then 9minutes and so on. The smaller the demand the longer the off period.
This means that it is highly unlikely that the network will be able to operate as a satisfactory replacement for a dedicated store. A heat pump needs to be off for 15minutes to allow the motor windings to cool before restart. Our 6m3 network cannot provide for this without significant use of a backup gas boiler.
In contrast a dedicated thermal store of 50m3 provides a minimum of 30minutes of continuous run time and means that the heat pump could, at say a load of 500kW allow 2 hours of off time – as might be typical for a summer day as shown below. It is not simply the greater volume that allows this but also the larger delta T available to the store. In this case 1m3 of the water in the store provides 4x the heat storage capability of 1m3 of water in the network. And the network delta T is peanuts here. A delta T of 40C is typical in Denmark. So the same volume of water in a store would give 8x the storage ability of our network deadband in a well-designed network.
The Red electricity tariff period occurs between 16:00 and 19:00 in the London Power Network area – slightly different times apply around the country. The 50m3 store means that the heat pump can avoid operation for 2 hours of this period without the use of gas boilers. An even larger store of 75m3 would mean for the same conditions the entire 3 hours could be avoided. An even larger store would increase the load conditions for which this applies – as well as providing greater flexibility to capture other grid incentives.
My example here is of course a relatively small network but in a larger network you’d have larger supply plant and be comparing against a larger dedicated thermal store. As we get more and more intermittent generation on the electricity then there will be more and more variation in halfhourly electricity prices, incentivising more flexibility. In Demark they regularly seeing negative electricity prices – you’re paid to consume excess wind generated electricity. So my message here is don’t rely on the network as a store – put in a dedicated store every time.

This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *