Fuelstretcher” Boiler Control Trial
HMNB CLYDE
Background
Even with the most highly sophisticated building management system installed and working to its full potential, there is still an efficiency gap in the control strategies of building heating systems which can result in overheated buildings. Boilers, in essence, require to supply the correct degree of heat energy to balance heat losses and gains, in line with weather conditions, in order to maintain internal temperatures.
To this end, there are two main contributors to heat wastage in buildings;
For good energy control, a buildings heat “in-flows” require to be dynamically reconciled with the “out-flows”, so as its specific heat profile can be matched. This concept is the cornerstone of the Fuelstretcher boiler control system
Fuelstretcher Boiler Controller
Essentially, Fuelstretcher is an intelligent, self learning boiler control system which learns user patterns and adjusts boiler operation, and hence fuel consumption, based on demand.
Existing control systems, at best, make use of inside and outside sensor temperature networking to drive controls to achieve internal comfort conditions. These control systems, correctly programmed, generally make good comfort decisions, however they are not always the best from an energy efficiency stand point.
What the Fuelstretcher attempts is to impose a higher order of intelligence in terms of efficient and dynamic control with regards to boiler control. It analyses the heat profile, via the primary flow and return pipe work, then predicts and manages the output of the boilers to match it in a continuous learning process, on a day to day basis. In turn, this results in the optimum primary temperatures for building comfort demand, set by the other building controls, and boiler energy conservation, therefore meaning that no excess heat is required to be absorbed by the heating system, the building fabric or the space itself due to overshoot of set points
Scope
An MTHW boilerhouse, feeding a general workshop and a motor transport workshop was seen as a good candidate for trialling the Fuelstretcher boiler controller with a view to reducing fuel bills as a result of better primary loop control. The reasons for choosing this boiler house was that, at the time the trial was undertaken, the building was demanding heat 24/7, and therefore any savings made to the rather substantial fuel bill could be directly apportioned to the controller rather than any other factor. In trialling this building it would also serve as a marker as to how successful the controller could be in providing savings for a boiler house supplying 24 hour serviced buildings, as are prominent on site.
Concerns
As with many “black box” retrofit controls promising instant savings, there is understandably always a degree of scepticism, and this is no different. Therefore the trial will have to address the following major doubts;
Fuelstretcher Trial
The Fuelstretcher was temporarily installed on the 16th March 2007, assuming control from the existing Landis & Gyr boiler sequencer.
Prior to this, from December 2006, weekly fuel consumption and degree days, (to a base temperature of 15.5°C), were recorded from the site BEMS so as to provide a sound datum, with the same information being collated after installation.
It was imperative to the trial that fuel use was weather compensated, and equally so that the information was sound and irrefutable. Therefore, due to the unusually high internal set temperatures within the dependant buildings, a standard base temperature of 15.5°C would not be representative of the buildings thermal performance, and 20°C was deemed a more appropriate base temperature. To that end, actual degree day data to a base temperature of 20°C for the West of Scotland was purchased to cover the trial period so as more representative degree day figure could be utilised for the trial period.
To further explain the rationale behind this, as it has a major bearing on the end savings results, normally, degree days use a base temperature of 15.5°C to maintain comfort levels in the region of 19 - 20°C, internal heat gains making up the difference. A standard deviation to this is for hospitals where a base temperature of 18.5°C is used where internal temperatures are required around 22 - 23°C. However, as one of the areas within the general workshop was attempting to maintain an internal temperature of 23°C with limited internal gain, 20°C seemed a reasonable base figure. (This base figure assumption was infact later confirmed by calculation to be a true representation).
Table A.1 shows the weekly fuel consumption and degree days for the two periods
Datum – Original Control |
Trial – Fuelstretcher Control |
||||
|
|
||||
Week Com |
Weekly D/d (base 20°C(1)) |
Fuel (kWh) |
Week Com |
Weekly D/d (base 20°C(1)) |
Fuel (kWh) |
|
|
|
|
|
|
25/12/06 |
103 |
34,715 |
19/03/07 |
82 |
30,263 |
01/01/07 |
102 |
39,008 |
26/03/07 |
109 |
30,104 |
08/01/07 |
85 |
38,743 |
02/04/07 |
84 |
25,758 |
15/01/07 |
93 |
34,397 |
09/04/07 |
84 |
21,412 |
22/01/07 |
102 |
34,556 |
16/04/07 |
59 |
25,652 |
29/01/07 |
111 |
43,248 |
23/04/07 |
70 |
21,359 |
05/02/07 |
92 |
30,157 |
|
|
|
12/02/07 |
128 |
43,354 |
|
|
|
19/02/07 |
94 |
43,248 |
|
|
|
26/02/07 |
81 |
39,951 |
|
|
|
05/03/07 |
101 |
34,291 |
|
|
|
12/03/07 |
89 |
38,902 |
|
|
|
|
|
|
|
|
|
Totals |
1,181 |
454,570 |
Totals |
488 |
154,548 |
Table A.1
Savings
From the two totals therefore, the average energy consumption corrected for temperature may be calculated for the original controller and the Fuelstretcher trial.
(NB. The Fuelstretcher trial period, by necessity, was shorter than the datum due to the fact that the heating season was coming to an end).
Original Landis & Gyr Control;
Ave. Energy Use / Degree Day = Weekly Energy Consumption Total
Weekly Degree Days Total
= 454,570
1,179
= 385 kWh / Deg d20
Fuelstretcher Control;
Ave. Energy Use / Degree Day = Weekly Energy Consumption Total
Weekly Degree Days Total
= 154,548
488
= 317 kWh / Deg d20
Therefore the energy saving achieved with the boilers under Fuelstretcher control during the trial period can be quantified as;
Average Energy Saving = (385 – 317) x 100
385
= 17.7 %
In monitory terms, using the average annual energy consumption figure from 2004-2006, of 2,462,818 kWh, and a gas oil cost of £0.03274 per kWh, (34.7p per litre), an indication of the actual cost savings can be evaluated;
Saving = (2,462,818 x 0.177) x 0.03274
= £14,272
Internal Comfort Levels
Figures A.1 and A.2 show temperature graphs representative of the internal conditions for the general workshop, both using the original control and the Fuelstretcher;
General Workshop Building - Original Boiler Control
Fig. A.1
General Workshop Building - Fuelstretcher Control
Fig. A.2
As can be seen the comfort levels are broadly comparable, although it was difficult to define a pattern from any of the temperature graphs for this building.
It would appear in the Contain Bay, and to a lesser extent the Machine Shop, that following a sharp temperature dip, which is undoubtedly when the vehicle doors are opened, there is an over shoot, possibly indicating the time for the control valves to modulate closed after re-heating, but to put this overshoot in to context, it averages less than 1°C. It is also evident that during milder ambients the internal conditions follow the outside temperatures, being a degree or so higher in these spaces during these times, however the same is not true during colder ambients when troughs are not evident.
The Inspect Office temperature profiles shows the largest variations, which again appear to have significant temperature rises during warmer ambients, most probably due to solar gain, but settling out at approximately the base temperature for most other times, with any dips probably being attributable to the occupants opening windows as a means of temperature control.
The MT workshop, as can be seen in Figures A.3 and A.4, shows a regular internal temperature pattern both before and after the Fuelstretcher was installed. The same sharp temperature drops and overshoots attributable to the use of the large vehicle doors are visible as were with general workshop, although the internal temperature changes during warm ambients are not evident.
MT Workshop Building – Original Control
Fig. A.3
MT Workshop Building – Fuelstretcher Control
Fig A.4
It is fair to say however that there has been no significant detriment to the building temperatures under the Fuelstretcher control, therefore the 17.7% fuel saving has been achieved without compromise to the building comfort levels.
Domestic Hot Water
Due to the lower primary temperature under Fuelstretcher control, it was to be expected that the LTHW temperature would also be lower, and therefore, although this would have no effect in the MT workshop due to it having a purely electric DHW calorifier, it could lower the general workshop`s storage calorifier primary temperature to an extent that it could not maintain its set temperature by LTHW alone, and would have to be assisted by its back up electric element. As can be seen in Figures A.5 and A.6, this was in fact the case.
General Workshop Domestic Hot Water Calorifier – Original Control
Fig. A.5
General Workshop Domestic Hot Water Calorifier – Fuelstretcher Control
Fig. A.6
As is shown in Fig. A.5, under the original control system the DHW is being overheated, whereas in Fig A.6 it is obvious that under Fuelstretcher control it is being under heated.
The energy consumption for the calorifier’s electric element over the trial period therefore requires to be quantified so as to give the true savings under the Fuelstretcher control. It is likely, however, that during the colder months the MTHW primary temperature chosen by the Fuelstretcher will be sufficiently high so as the DHW set point will be met via the LTHW circuit.
From graph Fig. A.6;
DHW Calorifier Electrical Element Rating = 24 kW
Average Weekday Element Use Over 24 Hours = 6 hours
Therefore, the additional annual electrical consumption for heating the DHW system is 52,560 kWh. This is a worst case scenario however which assumes that the electrical element is the sole heat source for the storage calorifier, and that the element operates for the same amount of time every day regardless of occupation. These assumptions are extremely unlikely and therefore the actual electrical consumption is likely to be somewhat less.
If this cost is calculated at an electricity charge of £0.062 per kWh, it equates to an additional energy cost of £2,444 annually, therefore the actual savings achieved by the Fuelstretcher can be given as;
Total Saving = Fuel Energy Savings – Additional Electrical Energy Costs = £14,272 - £2,444
= £11,828
N.B. This figure, as well as utilising the pessimistic electrical element usage figures, also takes no account of the wasted fuel energy previously used in overheating the hot water.
It should be noted that if this is unacceptable an occasional boost can be incorporated into the Fuelstretchers intelligence so as to ensure that the calorifier temperature may be maintained by its primary LTHW circuit, this will however, obviously have an effect on the fuel consumption.
(1) Sourced from Degree Days Direct Ltd