BLOG: Power Perspectives

[#1]

There is an age-old debate on what is better – a single, large generator set or paralleling multiple, smaller generator sets. There are good arguments on both sides.

Pro Paralleling:

• redundancy
• efficient operation through cycling gensets off-line based on load
• easier installation

Pro Single Unit:

• decreased complexity – no paralleling equipment need
• increased efficiency at higher loads
• simplified maintenance

Before the question can be truly answered, you need to consider specifics of you or your customer’s installation and requirements. There are a number of factors to consider, including:

• physical installation requirements
• basic site requirements – size of critical loads, load profile
• required reliability and redundancy requirements
• maintenance considerations
• owning & operating costs
• controls complexity

Suppose you have a facility with a load requirement of 2 MW. What is the best method to provide generator set power – a single 2 MW unit, two 1 MW units, four 500 kW units, forty 50 kW units? (I think we can rule the last one out.) Let’s take a look at one of the factors above – physical installation requirements. The larger generator sets typically have a higher power density. Not including working clearance around the generator sets, the total footprint of each option is roughly:
   1 x 2 MW genset: 164 sq. ft.
   2 x 1 MW genset: 182 sq. ft.
   4 x 500 kW genset: 180 sq ft.


And when working clearances are added, the single large generator set will take significantly less space than multiple, smaller generator sets. But the single, larger generator set cannot be broken up into smaller blocks, so you need to consider not only the total footprint but the footprint of the individual units. A 2 MW genset is 30% longer that a single 1MW genset, and 40% longer than a 500 kW genset. If there is a minimum length requirement driven by the site, smaller paralleled units may be a better fit for your site.

Of course, this does not consider the additional equipment required for installation – fuel systems, exhaust systems, distribution, and paralleling equipment – that would be required for multiple smaller generator sets. Or outdoor installations, where the generator set would be installed in an enclosure.

• How important are space constraints in your installation(s)?
• Do you have any examples where space played a role in your decision process?
• What are some other factors that you consider when determining if paralleling generator sets is right for your installation?
• What are some of the concerns you have with paralleling units? With a single unit?

 
I will be exploring this topic more in upcoming blogs, and I am interested in your opinion. Please post below.

[#2]

Multiple units paralleling would be technically feasible for industrial applications if we consider the following;

1. Lets take Steel Factory with  5MW max demand (approx 50 - 100 acre area) - Foundry is located far away from rolling mill, and pump station is remotely located, then the wire mill, power sub station would be remotely located. In this case multiple sets selection is really important as it is possible to install all these generator sets at their respective locations rather than one single large set of 5MW at the sub station.

As indicated above, it is possible to add and remove the generator sets in to the system, based on the average demand and hence it is possible to maintain minimum loading level of 30% of generator capacity for optimum performance of the system. This level of greater flexibility is essential for  industrial applications ,where their site load profile is variable and not flat.

Again,  by adding one at a time you can adjust your cash flow of this project and install the the generator sets one at a time and also, could achieve the greater flexibility of overall Project , and O&M management.

2. When you need to transport generators in to a remote site where road transportation is practically difficult you need to consider multiple small units rather than one big giant. Imagine a situation that you need to transport a generator set (black start) to hydropower station, or Off shore, applications, mining applications.

3. Lets  take an application of off shore platform where you need approximately 1MW diesel set. Then it is technically feasible to install 2 off 500kW sets as it is easy to install and will give an additional advantage of redundancy.

4. These days development of synch controller technology is impressive, and very competitive as Comap, Deepsea, Woodward easygen controllers are available in the market and could easily integrate with Diesel & Gas Generator control, and therefore, multiple units application would be financially viable in industrial applications, and could recover the cost of installations with in a short period of time.

5. Again,  in power station applications ie utilities, and embedded generator applications multiple installations gives greater flexibility of O&M , and project management of the installation.

6. If you consider the block loading capabilities and transient response, you need to ask the question that " will multiple generator installations would be a better option rather than one single large unit" ?

Ex: In Gas generators of 2MW range transient response to be studied and compared with mutiple unit option.Sudden 100% load removal cannot be applied in large single units.

7.On the other hand,  in CHP applications you need to do a feasibility study of installing multiple units,  and it would not be technically feasible as you need to hook up the  heat recovery systems to the each generator exhaust, Jacket water lines.

In this case,  for 2MW range,  it would be technically feasible to select one single unit of 2MW (G3520C) rather than multiple units. (continuous applications, and flat load profiles).

7. If you need to choose multiple generators for the  application you need to carefully study the following areas as well;

(I) Harmonics load profiles at the site VTHD requirements

(II) Neutral switching (one neutral at a time)

(III)Generator Data comparison ; ie  winding pitch, Zero seq reactances. (are those identical generators or different)

(IV) Additional protection requirements (ANSI 32, 25)

(V) Local regulations.

[#3]

The debate is simple...

Do you like to drive a vehicle without a spare tire?

Does it get any simpler than this

Gary Minker

Radio Works R.F. Consulting

[#4]

Agree. Everything narrow down to system reliability.

Decision depends on criticallity of loads served. Typically, an N+1 configuration is desired with the standby unit sized to provide at least 50% of the load. For system with really high demands busses and generators may required to be broken down is smaller distribution blocks to allow the power to be distributed safely.

In the example facility a 3 x 1 MW unit may be the minimal. However, if space is of a concern a 4 x 500 kW may prove to be a good choice and the one with higher operational flexibility. Load shedding schemes could also prove effective to improve up time.

If criticallity is not an issue 2 x 1 MW or even 1 x 2 MW can be considered. The first will be my option unless space is limited. For standby applications I will consider only one generator.

In any case a cost screening and risk analisys/comparison could help in the selection of the best alternative for a particular project. 

[#5]

Security issues require a new philosophy for paralleled applications of generator sets.

1. Paralleling should be a luxury not a necessity. Emergency power operation must not depend on a minimum number of generator sets serving the emergency power bus.

2. For true redundancy the smallest of the generator sets must be able to serve all of the critical load.

3. Paralleling generator sets must be separated by a two hour firewall.

4. Switchgear must be protected by a two hour firewall.

5. If the switchgear is equipped with hot swappable PLCs part of the acceptance test must be turning off power to the PLCs alternately.

6. If the switchgear is PLC controlled there must be a default manual operating condition proven at acceptance by turning off power to any and all PLCs.

7. If any PLC is part of a SCADA system or is accessible via the internet antivirus and hacking protection must be provided.

8. Paralleled generator sets must not be served by one common support system. This includes fuel, cooling, ventilation and fire suppression.

9. Switchgear control power must be provided by a best battery system plus a dedicated set of control power batteries.

10. Non critical monitoring must alarm for a reasonable time period before stopping important operations.

11. All control systems must be constructed to allow for the minimum PPE requirements.

12. No high energy devices should not share a room with the control system.

[#6]

Essa and GaryM are correct and in my opinion 100% to the mark. For reference, most of my career was spent serving the marine (including Navy), aerospace, large industrial and mining, and isolated utilities in a variety of Engineering roles. All of this experience and education said the bottom line considerations are the application, regulatory and/or technical society requirements, and the client's and professional prepared specifications - all of which embed 10's to hundreds of years of past design experience SPECIFIC FOR THE APPLICATION.   These key factors will determine this issue. Safety and mission criticality are additional factors. Many applications REQURE paralleling multiple power sources each with separate support systems; E.G. I cannot even imagine flying on a commercial aircraft or sailing on a surface combatant ship without these required design elements.

wlj1943

[#7]

Interesting debate.  I like Progen904's thinking, but cost will come into play at some point and some of those items seem like luxuries to me (2 hour fire rated walls separating everything for example).  It's also often not possible to keep the critical load less than the capacity of the smallest generator.  On larger systems that tends to be our biggest challenge.  We have several installations where we have two 1.5-2MW gensets up and closed inside of 10 seconds enabling us to close and accept priority 1 loads only at that time.  The more gens you have in the lineup, the better your odds of getting two closed in time.  To say that's a terrible gamble is and understatement, but necessity is the mother of invention.  Now, the advent of advanced and reduced cost controls allow us to put multiple tie breakers in the paralleling bus to allow single gensets to close to dedicated critical loads, THEN parallel across the ties and close in together.  Removes the stress on the equipmend and your blood pressure.

Those two points made, the decision to parallel then comes down to cost and space.  A single unit will (generally) cost less and occupy less space.  A parallel system gives you the ability to make decisions if a unit fails.  Redundancy is a wonderful thing. 

One fact that is often overlooked in this debate, however, is the intelligence of these parallel controls.  Many engineers feel that a parallel system provides them the redundancy they're looking for with load shed and load demand modes standard.  This is a fallacy.  The reality is most manufacturers offer these items as options only - and some not at all.  Want to see pain?  Stand in front of an IT Manager who thought he bought a system that would work even if a unit failed only to find out he lost a unit and in turn lost the whole plant.  Assume nothing.  Just as a plane with a single engine is an exercise in bravery, so too is buying equipment like this from the low bidder.  Caveat emptor...