Tuesday, 29 January 2013

Stirling Engine Blog 15 January 29 2013


I ran the engine today for 60 minutes using the new 800 W load. It appears to have worked exactly as planned - the unit's heater clamp did not come on even though the batteries were more or less fully charged at the start of the test. The coolant system reach equilibrium at around 42°C. The system reached steady state around 50 minutes into the test. 
Electric Heater used as battery load

Water condensing out of the exhaust gas (note a single droplet at the base of right angle fitting). The shape of the exhaust flue makes the fowl water drain ineffective and so the water leaks out of the seal. This is not an issue as it can be collected with a drip tray underneath. After about 30 minutes of runtime only about 20ml of water leaked.

Today's key data:
The system reaches a steady state at around 50 minutes into the test. The coolant system stays stabilises at around 42°C with the mains pressure at max flowrate. During efficiency tests, it will be necessary to let the coolant reach its setpoint temperature (60°C) to give a realistic efficiency measurement.

Monday, 28 January 2013

Stirling Engine Blog 14 January 28 2013

As mentioned in previous blog posts, the biggest unresolved problem was connecting the secondary heat exchanger to the mains water tap. It was necessary to join a standard garden hose to 3/4" qualpex - this is a very unusual combination and it is almost certainly not available as a standard fitting in a plumbing shop. I decided to quickly machine a small insert on my lathe that would allow me to join the flexible pipes. Whilst not very aesthetic the jubilee clip/rubber tubing has proven to be extremely effective and simple to use. 

Machining the part from a piece of scrap aluminium

Finished part

Preparing to join the two hoses

The garden hose fits tightly around the insert

The larger diameter hose also fits well. This larger hose just connects onto the qualpex.

Jubilee clips added for additional safety. The interference fit is probably sufficient

I replaced the fitting on the tap with a standard garden hose quick release

With the last of the plumbing problems solved, I hooked up the inverter and a small 250W heater load. I intend on replacing this with an 800W heater in the coming days to act as the proper load for the engine to slow the rate of battery charging.

Batteries and inverter setup


View of the test area (note MicroMon running on the computer)

All of the major systems work on the engine is now completed. There are only minor jobs remaining, mostly related to adding various sensors to the setup which will be logged by LabView.

These include:


  • Pressure Transducer - To view the pressure levels inside the engine and keep an eye on any leakage
  • Water Flow Meter - T measure the flow rate of mains water pressure through secondary heat exchanger
  • Water Flow Meter - To measure the flow rate inside the cooling loop (possibly unnecessary)
  • Multiple Thermocouples - To measure heat rejected to cooling circuit and to mains water dump


I also tested the engine with the new secondary heat exchanger dump. It significantly slowed the heating rate of the cooling system. It appeared to stabilise at around 40°C even with around 10 amps going into the heater clamp (the batteries had switched to absorption charge). I ran the engine for a total of about 30 minutes and logged the run using MicroMon.

Today's data. Tblock (blue) is a proxy for the cooling system temperature as both were in rough equilibrium. The key thing to notice is that even with the clamp heater activated, the secondary heat exhanger is able to dump enough heat out of the system to maintain more or less constant coolant temperature. 

Saturday, 26 January 2013

Stirling Engine Blog 13 January 26 2013


I did not have time to do any physical work on the engine today. I did however manage to take a look at some of the data taken from Thursday's test run and graph it (shown below).

There are still a number of tasks that remain to be completed before the engine is fully operational:
  • Connect Inverter (the battery clips need to be replaced with rings)
  • Machine an adapter fitting for the tap (garden hose to 3/4" qualpex)
  • Install Labview of the cell computer
  • Get a suitably sized load for the inverter (~800 W)

Preliminary view of some key engine parameters (Exhaust Temperature, Engine Block Temperature, Power Output). Note that the data logging was only switched on part way through the engine test and does not have show the engine startup behaviour.

Stirling Engine Blog 12 January 24 2013


I made a good deal of progress today with the engine. I decided to bring in my own drill in order to do some trivial tasks that were taking too long for to be done by others. I mounted the header tank board to the diesel engine frame and I also mounted the coolant pump to the wall. The header tanks are now secure on top of the diesel engine frame and reasonably well isolated from the Stirling engine. 

Exhaust duct in place and header tanks filled and mounted on the secured board

I also spent a good deal of time connecting the new radiator fittings and reinstating the radiator and associated plumbing. As mentioned in the previous blog, I incorporated a drain valve into the system which greatly simplifies draining the whole system.

New valves on the radiator. Note the drain valve on the right hand line of the radiator

After troubleshooting all of the plumbing and bleeding both the coolant and fuel lines of air I proceeded to test fire the engine for a short 10 minute run. It is not possible to run for any longer than this as the heat dump quickly saturates. The engine ran smoothly with no errors. I logged a portion of this data which will be discussed in a later blog.




Stirling Engine Blog 11 January 23 2013


I put the exhaust duct up today, it was relatively straightforward as the flexible duct fits perfectly over the exhaust flue. I am still waiting on the aluminized tape to make a proper seal at the base near the engine.
During the first test charge of the plumbing system (with just water and no glycol), I had a major leak in one of the radiator fittings which forced me to drain the whole system. The radiator had 2 compression fittings connected to it (jury-rigged). The fittings were not the correct fittings for the radiator and were only a temporary measure during the initial tests at home. The thread on the connecting nut appears to be stripped or otherwise damaged. It was necessary to go back to Heat Merchants and pick up some new radiator valves. I've decided to put in the proper radiator valves and use 1/2" to 3/4" straight compression fittings to make them compatible with the qualpex. I've also decided to incorporate a drain valve in the system to make draining the system easier.

Radiator disconnected from circuit

Radiator fitting that needed to be replaced. The thread on the silver hex ring is damaged and it was necessary to replace the whole fitting (this also required getting a radiator spanner - a 1/2" allen key essentially)

Tuesday, 22 January 2013

Stirling Engine Blog 10 January 22 2013


Pat did an excellent job of making an exhaust pipe. He threaded some thick tubing using a lathe and then welded the joint making it into a single piece. He then gave the tube another pass on the lathe to clean it up so that the whole arrangement fits flush inside of the flexible ducting. I plan to fit the exhaust and duct tomorrow and will use some aluminized tape to seal the joint near the engine exhaust.

Welded gunbarrel fitting with fabricated tubing

Exhaust pipe in place

I connected the fuel line and pump also. I decided to replace the connections on the fuel pump as they were frayed and loose. I used some barrel crimps and positioned them so that polarity reversal wouldn't be possible.

Blue barrel crimps on the fuel pump connections

In addition to the exhaust, I rerouted some of the plumbing to tidy it up but there is still some more work to be done before the rig is in an acceptable condition. Pat also made a stand for the header tanks out of a sheet of plywood. The tanks sit flat on the sheet with their outlets coming out through holes in the sheeting.

Rerouted plumbing (some work still needed). Note white fuel line and coolant header tank (extreme right corner)

All that remains to be done:
  • Get the inverter installed and hooked up
  • Mount the coolant pump to the wall
  • Connect the cold water pipe to the tap
  • Charge and bleed the coolant system
  • Charge and bleed the fuel system

Monday, 21 January 2013

Stirling Engine Blog 9 January 21, 2013


Pat Donnellan completed the risk assessment today and will prepare the associated paperwork with recommendations. He mentioned that he does not think there will be any major issues that need to be dealt with.

I also completed a "Node Registration Form" for ISS which is required to get the cell computer on the network. Once again, it is not clear what the lead time on this will be. 

There were no technicians on available today to discuss where I could put the header tanks for the engine so physical work on the unit is once again at a standstill. It is hoped that the exhaust fittings will be machined tomorrow and the exhaust can be installed and connected to the cell extraction system.

I purchased a 1000 Watt inverter over the weekend to allow me to connect a standard 240V AC mains load and discharge the battery bank when the engine is generating power. This will prevent the engine from ramping down and dumping current into the clamp heater when the batteries exceed a certain threshold (80% charge).


1000W (24V) Modified Sine Wave Inverter  

Inverter datasheet can be found here (model no. 651.667)

Saturday, 19 January 2013

Stirling Engine Blog 8 January 18, 2013


A number of significant developments occured today. I was contacted by a former Whispertech employee via the Google blog who was kind enough to share a copy of the MicroMon software with me but more importantly a complete set of instructions accompanied it. I have now discovered what was preventing me from using maintenance mode in the program. A somewhat convoluted set of operations is needed during the install in order to get full functionality from the software.

From the manual:

To access the program, a shortcut to the ".exe" file should be created for the desktop. Firstly, open the folder "C:\Micromon_Release_1_0" and right-click on the file called "MM-Release.exe"; this will bring up a menu where "Create Shortcut" should be selected. Move the newly created shortcut to the desktop by left-clicking the mouse over the shortcut icon and dragging it over to the desktop. Next, right-click on the shortcut and select "Properties" from the menu. In the text box called "Target", the text should read:
C:\Micromon_Release_1_0\MM-Release.exe
In order for the application to work correctly, "-Maintenance-Mode" should be appended to the above. The text should now read:
C:\Micromon_Release_1_0\MM-Release.exe -Maintenance-Mode
[Note that there is a space between the last "e" in "exe" and the hyphen ("-") before Maintenance].

I am now able to utilize the full functionality of MicroMon and have the ability to log and set unit variables. With this significant breakthrough I can now begin building the LabView program to centralize all of the data acquisition.


I made yet another trip to Heat Merchants and DPL to get the remaining plumbing supplies. I've decided to use qualpex and compression fittings for the cooling system connections due to its simplicity. As with the initial testing at home, I am going to use the Thermo King hose and jubilee clips for any remaining joints. I still have to find a way to join the secondary cooling loop to the test cell mains tap. 




Compression fittings installed on the engine

Initial plumbing arrangement. The qualpex will need to be cut to size once the header tank is positioned.

Cooling circuit

Stirling Engine Blog 7 January 17, 2013


Limited progress was made today. I worked on getting network access in the test cell. This involves contacting ISS with the room RJ port number and the test computer's unique identifier (physical address). Both of these pieces of information were forwarded to ISS but there was no indication of lead time on the request. 

RJ port : CR2-050
Physical Address : 00-40-63-E4-EF-20

I met with William Kelly, head of facilities which resulted in some new administrative issues that need to be dealt with before the engine can be commissioned. William highlighted that a risk assessment needed to be done on the project before any further work could be completed. 

William mentioned in particular:

Carbon Monoxide - He mentioned that they have had problems with this in the past.
Diesel handling - He mentioned the risks associated with dermatitis induced by diesel skin contact. 

Mitigating these risks is not going to be an issue. A maximum of 5 litres of diesel will be stored in the engine header tank at any given time and the tank can be filled using gloves with suitable equipment (jerry can with nozzle/hose). Carbon monoxide levels will be negligible at most as the engine utilizes a diesel burner which burns the fuel almost to completion (Nicholas Farra's thesis estimates 99.99%) and any exhaust will be extracted using the exhaust extraction system. As a final layer of protection the cell has multiple toxic gas detectors in case of any build up of CO.
It is worth noting that the test cell was specifically designed for running diesel engines. 

A detailed risk assessment will be undertaken at 12pm on Monday with myself, Rory Monaghan and Pat Donnellan (engineering building safety officer) and any recommendation will be acted on.

Wednesday, 16 January 2013

Stirling Engine Blog 6 January 16, 2013


The desk finally arrived (courtesy of Pat Kelly). I have set up the various bit of IT hardware and I've organized all of the engine wiring so that it is now neatly coming out the back and allows the unit to be operated with the enclosure in place. I think this is important as I do not want people interfering with the engine inadvertently or otherwise. The engine is now connected to its battery bank and the microprocessor once again has power which will enable me to troubleshoot the MicroMon problems discussed previously.

There are still a number of outstanding issues to be sorted out:
  • Engine exhaust - I have bought a 90° gun barrel bend to connect to the exhaust. Pat Kelly says that he can thread a bit of tubing to connect to it. The exhaust ducting can then be placed over this and will be sufficient for removing exhaust fumes.   
  • Coolant loop - I need to follow up with the plumbers on the status of the engine's plumbing arrangements. I will most likely go ahead with my own plumbing if this will not be in place by Friday.
  • Header tanks - Both the coolant and diesel header tanks will need to be mounted on the wall above the engine. I will need to talk to Pat about the best way to go about doing this.
  • Internet Connection - I need to get a longer CAT5 cable to connect to the test cell RJ ports. I may also need to get the ports activated for outside connections.
  • Computer Software - I need to install Labview on the server 

Pictures of today's progress are shown below:

View of the test cell before organization

View of the engine wired up


Desk arrangement (note I moved it into the corner as it's neater)

Stirling Engine Blog 5 January 15, 2013


There was no physical work done on the engine  today as I am still waiting on Pat Kelly to deliver the desk to the test cell. Nonetheless I spent a good portion of the day studying some more documentation. I managed to read through two theses from the University of Toronto, both which were based on using the Whispergen PPS16 unit. While the core focus of both papers was not relevant there was still a good deal of useful experimental and hardware related information.  

Efficiency and Emissions study of a Redisdential Micro-Cogeneration System Based on a Stirling Engine and Fuelled by Diesel and Ethanol - Nicolas Farra (Master's Thesis)

Performance Analysis of a Stirling Engine Fuelled by Diesel and Ethanol and Transition to Bio-oil - Charles Habbaky and Adrian Boangiu (BSc Thesis)


I also discovered another interesting piece of market intelligence (conducted by PikeResearch) but sadly it is not publicly available. I intend on doing another trawl on the web using new keywords to see if there are any more recent market/technology intelligence reports related to Stirling cycle combined heat and power technology.

The report:

Information on PikeResearch (from their website):
Pike Research, a part of Navigant Consulting's global Energy Practice, is a market research and consulting team that provides in-depth analysis of global clean technology markets. The team's research methodology combines supply-side industry analysis, end-user primary research and demand assessment, and deep examination of technology trends to provide a comprehensive view of these industry sectors.
Emerging technology markets are inherently rife with uncertainty and risk. Pike Research provides its clients with market intelligence designed to clarify investment and expansion opportunities for industry participants. Applications for the group's research and analysis include:
  • Market sizing and forecasting
  • Market segmentation by geography, technology, customer type, end-user market, etc.
  • Analyzing the structure of emerging value chains
  • Insights on the timing of key market and technology transitions
  • Competitive intelligence
  • Assessing market demand for new products and assisting with go-to-market strategies
  • Providing inputs for business planning, budgeting, and forecasting
  • Benchmarking companies within the competitive environment

An interesting site that I came across that I need to investigate more thoroughly:

In addition to reading the theses I decided to email many of the major Stirling engine manufacturers requesting information and advice in relation finding out the current state of the industry. Due to the difficulty in accessing up to date market intelligence (they are all secured behind immense paywalls) this was the only remaining course of action. I do not expect much from this but it was worth a try.

Companies contacted:
  • Sunpower
  • Microgen
  • Infinia
  • Stirling BioPower
  • Efficient Home Energy (Whispertech successor)

Stirling Engine Blog 4 January 14, 2013


I met with Pat Kelly and two of the Engineering department plumbers today to discuss what work would be needed to get the engine commissioned. They are going to make up the required plumbing and they also agreed that it is a relatively simple job to get it commissioned. Pat is also working on getting a desk for me and dealing with the exhaust ducting. I hope that this will be done soon as it will allow me to begin working on some of the other aspects of the system for example the data acquisition. 

The rest of the day was spent reading the remainder of the EPRI literature and communicating with Umer Khan in relation to the MicroMon installation bugs. It will not be possible to troubleshoot this problem until I have the engine IT hardware set up in the test cell - dependent on getting a desk. Umer also kindly offered to use remote access software to view the problem himself, again this is dependent on getting the apparatus set up. I intend on keeping in regular contact with the technical staff in Engineering to ensure good progress is being made.

I also came across a very recent document (June 2012) reviewing the current state of the art in Stirling engines for CHP which lists the current players that have or are soon to release commercial engines. The document can be found here below and some of the key tables have been included below for brevity. 

A Review of Stirling Engine Technologies applied to micro-Cogeneration Systems





I also came across some other interesting papers and documents on the state of Stirling engine technology:

Technological development in the Stirling cycle engines (2008)

A review of solar-powered Stirling engines and low temperature differential Stirling engines (2003)

DFE2008 Residential Micro-cogeneration (Wiki)

Residential Cogeneration Systems: A Review of The Current Technologies (2005)

Stirling Engine Blog 3 January 13, 2013


I transported the engine, along with all of its subsystems into the New Engineering Building today. The engine will be housed and operated in the Diesel Engine Test Cell of the Thermodynamics Lab. 

Test Cell Location:



I hope to get the engine preliminarily operational early next week and then begin optimizing the test rig and defining the required variables to be measured. 

Stirling Engine Blog 2 January 11, 2013


I met with Pat Kelly (mechanical engineering chief technician) today in the New Engineering Building today and discussed transporting the engine into the college this weekend. There was no issue with me transporting the unit into the Thermodynamics test cell on Sunday and we agreed to work on getting it operational next week as per university safety requirements.

We also briefly discussed the following technical issues that remain unresolved:
  • Cooling the engine - The most likely solution will be to use the marine heat exchange and pass water through it in an open loop
  • Exhaust duct - Pat mentioned that an adapter fitting would need to be made and that it wouldn't be an issue
  • Table for IT hardware and table for the engine - Pat will find something 
Two new serial communication adapters also arrived in the post this morning and I hope to test them both out this evening and establish a reliable connection with the microcontroller. As a safety precaution, the second connector is a direct RS-485 to USB converter which will hopefully work for certain. 

Stirling Engine Blog 1 January 10, 2013


All of today was spent reading through the excellent EPRI (Electric Power Research Institute) report on the state of Stirling engines in 2002. While it is somewhat dated it provides an excellent snapshot of the state of the industry and identifies the dominant trends. Unfortunately more recent information is restricted to fee paying members due to its highly sensitive nature and the depth of investigation undertaken. The price of individual reports is very high and beyond the reach of this project (ranging from $950 to at least $75,000). Thankfully there does not appear to be any other reports solely dedicated to Stirling engines. Nonetheless the various residential CHP reports would certainly mention the status of the technology.

The report itself can be found here (http://www.engr.colostate.edu/~marchese/mech337-10/epri.pdf)  

Information on the organisation (from the website)

The Electric Power Research Institute, Inc. conducts research, development and demonstration (RD&D) relating to the generation, delivery and use of electricity for the benefit of the public. An independent, nonprofit organization, we bring together scientists and engineers as well as experts from academia and the industry to help address challenges in electricity.
[...]
The advisory process helps shape our annual research portfolio and the specific areas in which we conduct our work. Due to sensitive issues involving technological, business and economic decisions, much of our work is initially limited to paying members, though all work eventually becomes publicly available and free of charge.

An additional report that is unavailable to the public is shown below - it is worth trying to obtain this as it will be the most current data on the state of CHP (including solar)



Stirling Engine Report 2 - IT Problems


Download link for Micromon MO:



Stirling Engine Progress Report 2 


Similar to the previous report, most of the work in relation to IT equipment and in this report was completed over the Christmas holidays.

As mentioned in the previous report, the Whispergen controller offers a secondary data output from the motherboard that can be logged on a computer using the proprietary Micromon MO software. The research undertaken by Thermo King did not use any of these features as they could not get a hold of the software. Whispertech, the engine manufacturers have since sold out to a Spanish conglomerate Mondragon and the marine variant I am using is discontinued, therefore there is no support available.  Through some extensive online searching I managed to make contact with Umer Khan, a former Masters student in the University of Toronto, Canada. His thesis was based on tests undertaken using a similar Whispergen marine unit and he was kind enough to share the software with me. Finding the instruction manual and installation guides was also a task in itself as they were not widely available online. Nonetheless, I managed to obtain several documents through some marine distributors.

The data is outputted from an RJ45 port on the motherboard as an RS485 signal. In order to read this signal it was necessary to purchase an RS485 to RS232 converter and wire a CAT5 cable into it. This was a difficult task to complete as there were no schematics available for the pin output from the motherboard. It was necessary to use an oscilloscope to determine which wires were carrying the RS-232 signal. 

The converter worked the first time but it appears to have been a dodgy unit as it did not work again. It is not possible to short-circuit or otherwise "blow up" RS-232/RS-485 as it is designed to be rugged. It was necessary to purchase a second one, which thankfully works consistently. As a safety precuation, a direct RS-485 to USB converter was also bought. 

IT Hardware that will be used in the test cell. Note: a spare CRT monitor will likely be used instead of the flat screen display.

Connector pinouts established via oscilloscope. 


 New RS-485 to RS-232 connector and backup RS-485 to USB connector.


Despite the success in achieving a connection with the engine microcontroller there appears to be a bug with the software program. I suspect that it is due to a corrupted installation as I am unable to access the "Maintenance Mode" which is critical for logging data and setting controller variables and parameters. With the help of Umer Khan, I hope to solve this issue shortly and do not expect it to be a major roadblock. Once I have access to the unit controller and the full functionality of Micromon, I will be able to log controller data at 1 second resolution. According to Umer, it is possible to read this data into a LabView program and have all experimental data in a single program. I intend on reproducing a system similar to this that will generate "polished" data that can simply be graphed in Excel.






Stirling Engine Report 1 - Hardware Testing


Stirling Engine Progress Report 1


The Stirling engine was shipped over from Thermo King's R&D Division in Minneapolis, USA and arrived in Thermo King Galway in early December. I then transported the engine from Thermo King to my home in Oranmore in order to clean it up and verify that it is fully operational in advance of the project. Much of the work undertaken in this report was done sporadically over the Christmas holidays.

Whispergen PPS16 Marine Stirling Engine in the boot of my car after collection from Thermo King. The unit weighs about 100kg which makes lifting and transporting it a challenge.

 
Condition of the engine as received from Thermo King. Much of the plumbing was modified as per requirements of Thermo King Minneapolis R&D Division most of the wiring was in a mess and disconnected.
Simply reconnecting the plumbing and electrics to the factory settings was a difficult and confusing job. The engine appears to have been stored without its cover and has literally been gathering dust over the years. There appears to be little or no corrosion of key parts.

The Stirling Engine is a pressurized system filled with nitrogen at 24 Bar. One of the worries of the prolonged storage was progressive leaking of the nitrogen working fluid and the possibility that the engine may need to be recharged with nitrogen. Obtaining pressurized pure nitrogen is a relatively straightforward task however gas regulators capable of discharging in excess of 10 Bar are very specialized and uncommon. The charging valve on the unit is a  1/4" NPT schrader valve.  The appropriate gas lines and valve fittings were borrowed from Thermo King along with a pressure gauge to check pressurization. Luckily the system appears to have maintained pressure and is not in need of a nitrogen refill.

Pressure gauge with hose lines.

James Nallen generously supplied me with two DAQ units. These will be used for monitoring any additional thermocouples that are required to monitor the system. Additionally, a pressure transducer to monitor engine nitrogen pressurization and fuel flow transducer may be incorporated in the measurement system. Luckily the Stirling engine has an advanced onboard microprocessor that is equiped with many sensors.


The Stirling engine's microprocessor is powered from the same 24 V battery bank that it discharges generated power into. Therefore, in order to test that the engine is functional it was necessary to obtain a pair of 12 V batteries. Thermo King kindly supplied me with some second-hand gel cell batteries. 

Pair of gel cells being charged using two charge controllers.

In order to test the unit it was necessary to build a small workbench in my workshop to accomodate easy access to the engine. Both front and back access would be required to reconnect the various electronics and sort out the plumbing. It was necessary to lift the engine out of its enclosure to clean it down and access hard-to-reach areas. 

Work area consists of a set of thick wooden laths with a sheet of plywood spanned between two benches. 

Engine in-situ. Note chief technician in the background (my Dad).

Engine being lifted out of its enclosure for cleaning.


Alternate view


Rear view of engine

View of the untangled electronics. The additional shunts and various external voltage measurement hardware was removed. This was being used by technicians in Thermo King Minneapolis. It is my intention to use the onboard sensors to perform this (Thermo King did not have access to the controller software at the time). For the purposes of unit testing, it is unnecessary. Many of the unit sensors are still disconnected and unidentified. It will be necessary to identify these using the unit schematics.


View of the unit connected to the battery bank. The microcontroller was powered up for the first time and the electronics appear functional.

 
Engine sensors fully connected, identified and labelled. Note Dad cleaning the fuel pump.

 
Make-shift primary cooling loop test. Note advanced header tank technology (white funnel!) The system is now just charged with water and bled of air. I have only charged it with water as a temporary measure to test pumps and subsystems. The actual working loop will be water with glycol and a corrosion inhibitor with a proper header tank. 

Overview of current temporary plumbed circuit to test the system. All that is left to be done is make a fuel delivery system and bleed it of air. A temporary exhaust system is also required.

Some high temperature exhaust ducting and some additional water hose, both courtesy of Thermo King.
  
The Engine is now operational and can be powered up for short periods of time (the tiny heat reservoir limits run time to about 10 minutes before the coolant begins to boil).